A novel communication system of high capacity

By leveraging Time-Limited systems to exploit FAT DOF, the proposed communication systems enhance channel capacity and meet 5G requirements, addressing the limitations of Band-Limited systems in spectral mask-constrained environments.

CA3138371CActive Publication Date: 2026-11-03MICHEL FATTOUCHE
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CA3138371
Authority / Receiving Office
CA · CA
Patent Type
Patents
Current Assignee / Owner
Priority Date
2019-06-07
Filing Date
2020-06-03
Publication Date
2026-11-03
Estimated Expiration
2040-06-03

AI Technical Summary

Technical Problem

Existing communication systems face challenges in achieving high communication rates while reducing complexity, cost, latency, bandwidth, and power consumption, particularly in systems constrained by spectral masks, where high-frequency components are often disregarded as not contributing significantly to channel capacity.

Method used

The proposed solution involves designing communication systems that utilize Time-Limited (TL) systems to leverage arbitrarily large Finite Access Time Degrees of Freedom (FAT DOF) to match spectral masks, thereby increasing channel capacity beyond conventional Band-Limited (BL) systems, without requiring expensive mm-wave bands.

Benefits of technology

This approach allows for substantial improvements in channel capacity, meeting future 5G requirements without the need for costly mm-wave transmissions, by effectively utilizing FAT DOF to create a medium SNR region, enhancing transmission rates while minimizing power consumption and cost.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

Two inventive contributions are made for improvement of communications systems. A first derives the channel capacity of a Time-Limited (TL) system across a communications channel contaminated by interference and by noise. The potential increase in channel capacity compared to current communications systems is due to the availability of an arbitrarily large number of Degrees of Freedom (DOF) with finite access Time (FAT) in a TL system. A second takes advantage of the theory established in the first objective to design novel systems, referred to as Mask-Matched TL systems with FAT DOF, or MTF systems for short. The disclosure shows several embodiments of MTF systems where it is possible to improve the capacity of current communications systems, without having to modify or alter their Power Spectral Density, merely by taking advantage of their existing but unexploited FAT DOF through the 3 MTF design steps introduced in this disclosure.
Need to check novelty before this filing date? Find Prior Art

Description

A NOVEL COMMUNICATION SYSTEM OF HIGH CAPACITY 1 FIELD OF THE INVENTION The present invention relates generally to the field of communications where it is desirable to communicate between a number of transmitting units (TUs) and receiving units (RUs): 1. at a relatively high communications rate, over a communications channel, while 2. reducing the complexity, cost, latency, bandwidth (BW) and power consumption of the TUs and RUs. The present invention relates to methods and apparatuses for communications over a varied number of physical media such as: satellite, radio, audio, video, microwave, milli-meter (mm) wave, telephone line, subscriber loop, optical fiber cable, coaxial cable, twisted pair, USB cable, wireless, laser, infrared, power-line, tropospheric, ionospheric, Very High Frequency (VHF), Ultra High Frequency (UHF), etc. The present invention relates to methods and apparatuses for communications using a varied number of examples of TUs and of RUs such as: a base station (BS), or an access point (AP), a satellite transponder, a cellular telephone, a mobile telephone, a PCS telephone, a wireline modem, a wireless modem, a power line modem, a WiFi station, Zigbee node, a Bluetooth radio, computers, tablets, PDAs, sensors, watches, an Internet-of-Things (IOT) device, a Wireless Sensor Network (WSN) device, etc. The present invention relates to a varied number of communications networks such as: 1. A centralized network with a downlink (DL) portion from BS / AP to device, and / or an uplink (UL) portion from device to BS / AP, where we have referred to a BS / AP as the infrastructure unit of the network, and to a device as a unit receiving / transmitting services (voice, data, video, etc.) from / to the infrastructure units. A device includes a cellular telephone, a mobile telephone, a PCS telephone, a wireline modem, a wireless modem, a WiFi station, a Zigbee node, a Bluetooth radio, etc. Each BS / AP and each device comprises a TU and / or an RU. 2. A distributed network where all TUs / RUs act either as a BS / AP or as a device, or both, including mesh networks, multi-hop networks, peer-to-peer networks etc. Each node comprises a TU and / or an RU. The present invention relates to a varied number of communications applications such as ones with: 1. an UL emphasis such as Wireless Sensor Networks (WSN), 2. a DL emphasis such as Downstreaming Video (DV), 3. a Multicasting emphasis such as in radio and wireless systems, 4. a Broadcasting emphasis such as in AM, FM, DAB and GPS / Glonass / Galileo systems, and 5. a peer-to-peer emphasis such as driverless cars. The present invention relates to a varied number of standards such as: 1. 3G standards, 4G standards (also known as Long Term Evolution (LTE)), 5G standards, etc., 2. WiFi (IEEE 802.11a, b, g, n, ax, ac, ad, etc.) standards, 3. Bluetooth and Zigbee standards, 4. LoRa standards, 5. UWB standards, etc., to name a few. The present invention relates to a varied number of communications networks with various ranges between each TU and its designated RU(s) such as: 1. Ultra long range communications including deep space systems, 2. long range communications including satellite, micro-wave links, LoRa and cellular systems, 3. medium range communications such as WiFi and laser systems, and 4. short range communications such as Zigbee and Bluetooth systems. The present invention relates to a varied number of communications networks with various types of TUs and RUs such as: 1. TUs comprising one transmitter (Tx), also referred to as single input (SI), 2. RUs comprising one receiver (Rx), also referred to as single output (SO), 3. TUs comprising multiple Txs, also referred to as multiple input (MI) (also known as Multi-User (MU)), and 4. RUs comprising multiple Rxs, also referred to as multiple output (MO). The present invention relates to a varied number of communications networks with various types of MI such as: 1. MI which can collaborate, and 2. MI which cannot collaborate, such as MU networks. Based on the above, this disclosure uses SISO, MISO, SIMO, MU-MISO and / or MU- MIMO communications links. 2 BACKGROUND OF THE INVENTION In many applications, it is desirable to communicate between a number of TUs and a number of RUs over a communications channel, in an efficient way where complexity, BW, power consumption, latency and cost of each of the TUs and / or RUs are reduced while the transmission rate and range between each TU and its designated RU(s) are increased. In some applications, the specified channel is constrained by a mask, while the cost of each of the TUs and RUs is mainly dictated by its complexity. Reducing the power consumption of each of the TUs and RUs, while increasing its transmission rate, <com:Math><semantics>Rb<annotation encoding="application / x-tex">R_b< / annotation>< / semantics>< / com:Math>, across a channel constrained by a mask of BW, <com:Math><semantics>𝒲m<annotation encoding="application / x-tex">\mathcal{W}_m< / annotation>< / semantics>< / com:Math>, can be usually considered a trade-off between power efficiency and bandwidth efficiency, <com:Math><semantics>η<annotation encoding="application / x-tex">\eta< / annotation>< / semantics>< / com:Math>. Power efficiency is usually measured in terms of (how small) the minimum average received Signal Power-to-Noise Power Ratio (SNR), <com:Math><semantics>SNR¯<annotation encoding="application / x-tex">\overline{SNR}< / annotation>< / semantics>< / com:Math>, that is required to achieve a specific performance (such as Bit Error Rate (BER)), while bandwidth efficiency, <com:Math><semantics>η<annotation encoding="application / x-tex">\eta< / annotation>< / semantics>< / com:Math>, is usually measured in terms of (how large) the number of bits that can be communicated per second, <com:Math><semantics>Rb<annotation encoding="application / x-tex">R_b< / annotation>< / semantics>< / com:Math>, per Hertz (bps / Hz) of mask BW, <com:Math><semantics>𝒲m<annotation encoding="application / x-tex">\mathcal{W}_m< / annotation>< / semantics>< / com:Math>, between each TU and its designated RU(s) across a communications channel. This trade-off is often captured by an all-encompassing figure of merit, the well-known channel capacity, <com:Math><semantics>𝒞<annotation encoding="application / x-tex">\mathcal{C}< / annotation>< / semantics>< / com:Math>, which provides an upper bound on <com:Math><semantics>Rb≜η×𝒲m<annotation encoding="application / x-tex">R_b \triangleq \eta \times \mathcal{W}_m< / annotation>< / semantics>< / com:Math> that can be achieved as a function of SNR. This disclosure extends the capacity, <com:Math><semantics>CBL<annotation encoding="application / x-tex">C_{BL}< / annotation>< / semantics>< / com:Math>, of a band-limited (BL) channel to include the capacity, <com:Math><semantics>𝒞TL<annotation encoding="application / x-tex">\mathcal{C}_{TL}< / annotation>< / semantics>< / com:Math>, of a Time-Limited (TL) channel. Wyner first studied the capacity, <com:Math><semantics>𝒞TL<annotation encoding="application / x-tex">\mathcal{C}_{TL}< / annotation>< / semantics>< / com:Math>, of a TL system in 1966, after constraining the system to be approximately Band-Limited (BL). When the system is instead constrained to be Root Mean Square (RMS) BL, reducing interference between input signals is accomplished by minimizing the RMS bandwidth of each signal. The solution of such a minimization was shown by Gabor to be one lobe of a sine wave. Over time, the notion that practical communication systems are approximately BL was replaced by the notion that they are indeed BL as far as the channel capacity, <com:Math><semantics>CTL<annotation encoding="application / x-tex">C_{TL}< / annotation>< / semantics>< / com:Math>, is concerned. This was because it was thought that the high frequency components, which exist in a TL system and which fall far below the noise floor when the system is constrained by a spectral mask, couldn't realistically contribute to the channel capacity, <com:Math><semantics>CTL<annotation encoding="application / x-tex">C_{TL}< / annotation>< / semantics>< / com:Math>, of the TL system. The disclosure shows that such components can indeed contribute significantly to <com:Math><semantics>𝒞TL<annotation encoding="application / x-tex">\mathcal{C}_{TL}< / annotation>< / semantics>< / com:Math>, based on the fact that they represent an arbitrarily large number of Finite Access Time (FAT) Degrees of Freedom (DOF) with the ability to offer a linear SNR contribution towards the capacity. This is in contrast with BL systems, which contain a finite number of FAT DOF, and consequently, can only offer a logarithmic SNR contribution towards the capacity. By definition, the time to access a FAT DOF is finite. More specifically, this disclosure introduces novel means and apparatuses for designing communications systems and techniques that are able to increase their channel capacity, <com:Math><semantics>CTL<annotation encoding="application / x-tex">C_{TL}< / annotation>< / semantics>< / com:Math>, compared to the capacity, <com:Math><semantics>CBL<annotation encoding="application / x-tex">C_{BL}< / annotation>< / semantics>< / com:Math>, of existing communications systems by transporting incremental information across the arbitrarily large number of FAT DOF that exist in a TL waveform. In contrast, existing systems are assumed to be BL and therefore are assumed to have a limited number of FAT DOF with a limited ability to carry incremental information. Moreover, existing communications systems attempt to comply with the constraint of a mask through the use of certain filters, either digital or analog, selected primarily as to not cause a significant loss in orthogonality between input signals, while restricting their signal BW as much as possible in the hope of maintaining their BL status. On the other hand, the current disclosure takes advantage of the existence of the arbitrarily large number of FAT DOF in a TL system, by matching parts of the FAT DOF in a TL system to the imposed mask. We refer to such a matching as a Mask- Matched TL method with FAT DOF, or the MTF method for short, and refer to the apparatus, which performs such a matching as the MTF apparatus. Examples of the imposed mask in MTF methods and apparatuses include current wireless standards, such as WiFi and LTE systems, as well as future 5<com:Sup>th< / com:Sup> Generation (5G) wireless standards, which promise to deliver more than 10Gbps download capacities. In order to meet the promised 5G download capacities, current systems require a channel with an estimated median BW of 3.5GHz. This large BW is only available across mm-wave bands (26GHz, 28GHz, 38GHz, and 60GHz). Such bands suffer from a high path loss and are not multipath rich. The high path loss restricts coverage to Line-of-Sight (LOS) coverage, while a poor multipath environment limits the number of spatial DOF in a MIMO system. Moreover, mm-wave bands are expensive and potentially harmful. By relying on MTF systems, this disclosure shows how to exceed current wireless standards as well as how to meet and exceed the 10Gbps requirement for 5G systems with a BW, <com:Math><semantics>𝒲m=20MHz<annotation encoding="application / x-tex">\mathcal{W}_m = 20MHz< / annotation>< / semantics>< / com:Math>, with a carrier frequency in the mid-band spectrum of <com:Math><semantics>[2GHz,6GHz]<annotation encoding="application / x-tex">[2GHz, 6GHz]< / annotation>< / semantics>< / com:Math>, i.e. without requiring mm-wave transmissions. 3 SUMMARY OF THE INVENTION It is an objective of the current disclosure to design novel communications systems that are capable of offering substantial improvement in channel capacity compared to current communications systems. To this end, the disclosure derives the channel capacity, <com:Math><semantics>𝒞TL<annotation encoding="application / x-tex">\mathcal{C}_{TL}< / annotation>< / semantics>< / com:Math>, of a Time-Limited (TL) system across a communication channel contaminated by interference and by noise. The potential increase in channel capacity is shown to be due to the availability of an arbitrarily large number of FAT DOF in a TL system. Such FAT DOF are able to carry incremental information compared to the information that is carried by existing systems. By taking advantage of the arbitrarily large number of FAT DOF in a TL system, the disclosure shows that such incremental information can be made arbitrarily large as well, forcing the channel capacity, <com:Math><semantics>CTL<annotation encoding="application / x-tex">C_{TL}< / annotation>< / semantics>< / com:Math>, of the TL system to contain a new SNR region, referred to as the medium SNR region, in addition to the traditional low and high SNR regions found in the capacity, <com:Math><semantics>𝒞BL<annotation encoding="application / x-tex">\mathcal{C}_{BL}< / annotation>< / semantics>< / com:Math>, of a BL system. The newly created SNR region allows for the design of a novel TL system, namely the MTF system, where doubling <com:Math><semantics>𝒞TL<annotation encoding="application / x-tex">\mathcal{C}_{TL}< / annotation>< / semantics>< / com:Math> requires only a fixed multiple increase in SNR, as opposed to a BL system where doubling <com:Math><semantics>𝒞BL<annotation encoding="application / x-tex">\mathcal{C}_{BL}< / annotation>< / semantics>< / com:Math> requires a geometric multiple increase in SNR. The medium SNR region loads between 1 and 2 bits of information / DOF. This is in contrast with either the low SNR region, which loads < 1 bit of information / DOF, or the high SNR region, which typically loads <com:Math><semantics>>1<annotation encoding="application / x-tex">> 1< / annotation>< / semantics>< / com:Math> bit of information / DOF. Recently, several systems found a way to increase the number of their DOF. Such systems include MU systems, which form the basis for 3G wireless systems, and MIMO systems, which are currently adopted in most cellular standards, including 4G and 5G wireless systems. MU systems correspond to having K co-located users, each with a spreading gain N, which is the number of their DOF, while MIMO systems correspond to having K transmit antennas and N receive antennas where the number of DOF <com:Math><semantics>≤<annotation encoding="application / x-tex">\leq< / annotation>< / semantics>< / com:Math> min <com:Math><semantics>{K,N}<annotation encoding="application / x-tex">\{K, N\}< / annotation>< / semantics>< / com:Math>. Despite the fact that both systems have the ability to arbitrarily increase the number of their DOF, their respective capacities do not contain a medium SNR region since both systems fail to realize that, under certain conditions, some DOF, namely the FAT DOF, can offer a linear SNR contribution towards the capacity, by carrying incremental information. The current disclosure shows that in an MTF system, it is possible to force parts of the high SNR region to fall into the medium SNR region through the use of multiple receiving antennas such as in MIMO and / or in MU-MIMO communications. In summary, the disclosure shows several designs of MTF systems, which outperform current systems, including one design that is shown to meet future 5G capacity specifications without requiring mm-wave bands. 4 DETAILED DESCRIPTION OF THE DRAWINGS The present invention, both as to its organization and manner of operation, may best be understood by reference to the following descriptions, and the accompanying drawings of various embodiments wherein like reference numerals are used throughout the several views, and in which: • FIG. 1 is a schematic view of an embodiment of a SISO communications link between one TU 314 and one RU 315, over a communications channel 106, the link comprising o MTF Modulator 102, part of the digital side 510, which converts the (possibly Forward Error Correction (FEC) coded) information vector, <com:Math><semantics>α→<annotation encoding="application / x-tex">\vec{\alpha}< / annotation>< / semantics>< / com:Math> 101, into an MTF (digital) vector, <com:Math><semantics>β→<annotation encoding="application / x-tex">\vec{\beta}< / annotation>< / semantics>< / com:Math> 103, o transmitter, Tx 104, part of the digital side 510 and analog side 511, 513, which converts the MTF (digital) vector, <com:Math><semantics>β→<annotation encoding="application / x-tex">\vec{\beta}< / annotation>< / semantics>< / com:Math> 103, into a transmitted MTF (analog) signal, <com:Math><semantics>x(t)<annotation encoding="application / x-tex">x(t)< / annotation>< / semantics>< / com:Math> 105, o communications channel 106 which transports the transmitted MTF (analog) signal, <com:Math><semantics>x(t)<annotation encoding="application / x-tex">x(t)< / annotation>< / semantics>< / com:Math> 105 from Tx 104 to be received as <com:Math><semantics>y(t)<annotation encoding="application / x-tex">y(t)< / annotation>< / semantics>< / com:Math> 107 at Rx 108, o receiver Rx 108, part of the analog side 616, 618 and digital side 617, which converts the MTF (analog) signal, <com:Math><semantics>y(t)<annotation encoding="application / x-tex">y(t)< / annotation>< / semantics>< / com:Math> 107, into a received MTF (digital) vector, <com:Math><semantics>r→<annotation encoding="application / x-tex">\vec{r}< / annotation>< / semantics>< / com:Math> 109, and oMTF Detector 110 part of the digital side 617, which detects the (possibly FEC coded) information vector, <com:Math><semantics>α→<annotation encoding="application / x-tex">\vec{\alpha}< / annotation>< / semantics>< / com:Math> 101, as <com:Math><semantics>α→̂<annotation encoding="application / x-tex">\hat{\vec{\alpha}}< / annotation>< / semantics>< / com:Math> 111 from the received MTF (digital) vector, <com:Math><semantics>r→<annotation encoding="application / x-tex">\vec{r}< / annotation>< / semantics>< / com:Math> 109, using an estimate (possibly using a training sequence), which we refer to as Channel State Information at Receiver (CSIR) 115, of the state of the communications channel 106. • FIG. 2 is a schematic view of an embodiment of a MU-MISO communications link comprising K TUs 316, ..., 318, and one RU 317, across a Multidimensional communications channel 306, the TU 316 comprising o MTF Modulator<com:Sub>1< / com:Sub> 302, part of the digital side 510, which converts the (possibly FEC coded) information vector, <com:Math><semantics>α→1<annotation encoding="application / x-tex">\vec{\alpha}_1< / annotation>< / semantics>< / com:Math> 301, into an MTF (digital) vector, <com:Math><semantics>β→1<annotation encoding="application / x-tex">\vec{\beta}_1< / annotation>< / semantics>< / com:Math> 303, o and Tx<com:Sub>1< / com:Sub> 304, part of the digital side 510 and analog side 511, 513, which converts the MTF (digital) vector, <com:Math><semantics>β→1<annotation encoding="application / x-tex">\vec{\beta}_1< / annotation>< / semantics>< / com:Math> 303, into a transmitted MTF (analog) signal, <com:Math><semantics>x1(t)<annotation encoding="application / x-tex">x_1(t)< / annotation>< / semantics>< / com:Math> 305. The TU 318 comprises <com:Math><semantics>∘<annotation encoding="application / x-tex">\circ< / annotation>< / semantics>< / com:Math> MTF Modulator<com:Sub>K< / com:Sub> 202, part of the digital side 510, which converts the (possibly FEC coded) information vector, <com:Math><semantics>α→K<annotation encoding="application / x-tex">\vec{\alpha}_K< / annotation>< / semantics>< / com:Math> 201, into an MTF (digital) vector, <com:Math><semantics>β→K<annotation encoding="application / x-tex">\vec{\beta}_K< / annotation>< / semantics>< / com:Math> 203, part of the digital side 510, <com:Math><semantics>∘<annotation encoding="application / x-tex">\circ< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>TxK<annotation encoding="application / x-tex">Tx_K< / annotation>< / semantics>< / com:Math> 204, part of the digital side 510 and analog side 511, 513, which converts the MTF (digital) vector, <com:Math><semantics>β→K<annotation encoding="application / x-tex">\vec{\beta}_K< / annotation>< / semantics>< / com:Math> 203, into a transmitted MTF (analog) signal, <com:Math><semantics>xK(t)<annotation encoding="application / x-tex">x_K(t)< / annotation>< / semantics>< / com:Math> 205. The RU 317 comprises o Rx<com:Sub>1< / com:Sub> 308, part of the analog side 616, 618 and digital side 617, which converts the MTF (analog) signal, <com:Math><semantics>y1(t)<annotation encoding="application / x-tex">y_1(t)< / annotation>< / semantics>< / com:Math> 307, into a received MTF (digital) vector, <com:Math><semantics>r→1<annotation encoding="application / x-tex">\vec{r}_1< / annotation>< / semantics>< / com:Math> 309, and o MTF Detector<com:Sub>1< / com:Sub> 310, MTF Detector 110 part of the digital side 617, which detects the (possibly FEC coded) information vectors, <com:Math><semantics>α→1<annotation encoding="application / x-tex">\vec{\alpha}_1< / annotation>< / semantics>< / com:Math> 301, ..., <com:Math><semantics>α→K<annotation encoding="application / x-tex">\vec{\alpha}_K< / annotation>< / semantics>< / com:Math> 201 as <com:Math><semantics>α→̂1<annotation encoding="application / x-tex">\hat{\vec{\alpha}}_1< / annotation>< / semantics>< / com:Math> 311, ..., <com:Math><semantics>α→̂K<annotation encoding="application / x-tex">\hat{\vec{\alpha}}_K< / annotation>< / semantics>< / com:Math> 211 from the received MTF (digital) vector, <com:Math><semantics>r→1<annotation encoding="application / x-tex">\vec{r}_1< / annotation>< / semantics>< / com:Math> 309, using CSIR 116, which is an estimate (possibly using a training sequence) of the state of the Multidimensional communications channel 306. • FIG. 3 is a schematic view of an embodiment of a MU-MIMO communications link comprising K TUs 316, ..., 318, and <com:Math><semantics>Nr<annotation encoding="application / x-tex">N_r< / annotation>< / semantics>< / com:Math> non-collaborating RUs 317, ..., 319, across a Multidimensional communications channel 406, the TU 316 comprising o MTF Modulator<com:Sub>1< / com:Sub> 302, part of the digital side 510, which converts the (possibly FEC coded) information vector, <com:Math><semantics>α→1<annotation encoding="application / x-tex">\vec{\alpha}_1< / annotation>< / semantics>< / com:Math> 301, into an MTF (digital) vector, <com:Math><semantics>β→1<annotation encoding="application / x-tex">\vec{\beta}_1< / annotation>< / semantics>< / com:Math> 303, <com:Math><semantics>∘<annotation encoding="application / x-tex">\circ< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>Tx1<annotation encoding="application / x-tex">Tx_1< / annotation>< / semantics>< / com:Math> 304, part of the digital side 510 and analog side 511, 513, which converts the MTF (digital) vector, <com:Math><semantics>β→1<annotation encoding="application / x-tex">\vec{\beta}_1< / annotation>< / semantics>< / com:Math> 303, into a transmitted MTF (analog) signal, <com:Math><semantics>x1(t)<annotation encoding="application / x-tex">x_1(t)< / annotation>< / semantics>< / com:Math> 305. The TU 318 comprises <com:Math><semantics>∘<annotation encoding="application / x-tex">\circ< / annotation>< / semantics>< / com:Math> MTF Modulator<com:Sub>K< / com:Sub> 202, part of the digital side 510, which converts the (possibly FEC coded) information vector, <com:Math><semantics>α→K<annotation encoding="application / x-tex">\vec{\alpha}_K< / annotation>< / semantics>< / com:Math> 201, into an MTF (digital) vector, <com:Math><semantics>β→K<annotation encoding="application / x-tex">\vec{\beta}_K< / annotation>< / semantics>< / com:Math> 203, <com:Math><semantics>∘<annotation encoding="application / x-tex">\circ< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>TxK<annotation encoding="application / x-tex">Tx_K< / annotation>< / semantics>< / com:Math> 204, part of the digital side 510 and analog side 511, 513, which converts the MTF (digital) vector, <com:Math><semantics>β→K<annotation encoding="application / x-tex">\vec{\beta}_K< / annotation>< / semantics>< / com:Math> 203, into a transmitted MTF (analog) signal, <com:Math><semantics>xK(t)<annotation encoding="application / x-tex">x_K(t)< / annotation>< / semantics>< / com:Math> 205. The RU 317 comprises o Rx<com:Sub>1< / com:Sub> 308, part of the analog side 616, 618 and digital side 617, which converts the MTF (analog) signal, <com:Math><semantics>y1(t)<annotation encoding="application / x-tex">y_1(t)< / annotation>< / semantics>< / com:Math> 307, into a received MTF (digital) vector, <com:Math><semantics>r→1<annotation encoding="application / x-tex">\vec{r}_1< / annotation>< / semantics>< / com:Math> 309, and o MTF Detector<com:Sub>1< / com:Sub> 310, MTF Detector 110 part of the digital side 617, which detects the (possibly FEC coded) <com:Math><semantics>K<annotation encoding="application / x-tex">K< / annotation>< / semantics>< / com:Math> information vectors <com:Math><semantics>α→1<annotation encoding="application / x-tex">\vec{\alpha}_1< / annotation>< / semantics>< / com:Math> 301, ..., <com:Math><semantics>α→K<annotation encoding="application / x-tex">\vec{\alpha}_K< / annotation>< / semantics>< / com:Math> 201 as <com:Math><semantics>α→̂1<annotation encoding="application / x-tex">\hat{\vec{\alpha}}_1< / annotation>< / semantics>< / com:Math> 311, ..., <com:Math><semantics>α→̂K<annotation encoding="application / x-tex">\hat{\vec{\alpha}}_K< / annotation>< / semantics>< / com:Math> 211 from the received MTF (digital) vector, <com:Math><semantics>r→1<annotation encoding="application / x-tex">\vec{r}_1< / annotation>< / semantics>< / com:Math> 309, using CSIR 116, which is an estimate (possibly using a training sequence) of the state of the Multidimensional communications channel 406. The RU 319 comprises <com:Math><semantics>∘RxNr<annotation encoding="application / x-tex">\circ Rx_{N_r}< / annotation>< / semantics>< / com:Math> 208, part of the analog side 616, 618 and digital side 617, which converts the MTF (analog) signal, <com:Math><semantics>yNr(t)<annotation encoding="application / x-tex">y_{N_r}(t)< / annotation>< / semantics>< / com:Math> 207, into a received MTF (digital) vector, <com:Math><semantics>r→Nr<annotation encoding="application / x-tex">\vec{r}_{N_r}< / annotation>< / semantics>< / com:Math> 209, and <com:Math><semantics>∘<annotation encoding="application / x-tex">\circ< / annotation>< / semantics>< / com:Math> MTF Detector<com:Sub>N<com:Sub2>r< / com:Sub2>< / com:Sub> 210, MTF Detector 110 part of the digital side 617, which detects the (possibly FEC coded) information vectors <com:Math><semantics>α→1<annotation encoding="application / x-tex">\vec{\alpha}_1< / annotation>< / semantics>< / com:Math> 301, ..., <com:Math><semantics>α→K<annotation encoding="application / x-tex">\vec{\alpha}_K< / annotation>< / semantics>< / com:Math> 201 as <com:Math><semantics>α→̂1<annotation encoding="application / x-tex">\hat{\vec{\alpha}}_1< / annotation>< / semantics>< / com:Math> 311, ..., <com:Math><semantics>α→̂K<annotation encoding="application / x-tex">\hat{\vec{\alpha}}_K< / annotation>< / semantics>< / com:Math> 211, from the received MTF (digital) vector, <com:Math><semantics>r→Nr<annotation encoding="application / x-tex">\vec{r}_{N_r}< / annotation>< / semantics>< / com:Math> 209, using CSIR 117, which is an estimate (possibly using a training sequence) of the state of the communications channel 406. • FIG. 4 is a schematic view of an embodiment of a MU-MIMO communications link comprising K TUs 316, ..., 318, and one RU 321 with <com:Math><semantics>Nr<annotation encoding="application / x-tex">N_r< / annotation>< / semantics>< / com:Math> collaborating Rxs 308, ..., 208, across a Multidimensional communications channel 406, the TU 316 comprising o MTF Modulator<com:Sub>1< / com:Sub> 302, part of the digital side 510, which converts the (possibly FEC coded) information vector, <com:Math><semantics>α→1<annotation encoding="application / x-tex">\vec{\alpha}_1< / annotation>< / semantics>< / com:Math> 301, into an MTF (digital) vector, <com:Math><semantics>β→1<annotation encoding="application / x-tex">\vec{\beta}_1< / annotation>< / semantics>< / com:Math> 303, <com:Math><semantics>∘<annotation encoding="application / x-tex">\circ< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>Tx1<annotation encoding="application / x-tex">Tx_1< / annotation>< / semantics>< / com:Math> 304, part of the digital side 510 and analog side 511, 513, which converts the MTF (digital) vector, <com:Math><semantics>β→1<annotation encoding="application / x-tex">\vec{\beta}_1< / annotation>< / semantics>< / com:Math> 303, into a transmitted MTF (analog) signal, <com:Math><semantics>x1(t)<annotation encoding="application / x-tex">x_1(t)< / annotation>< / semantics>< / com:Math> 305. The TU 318 comprises <com:Math><semantics>∘<annotation encoding="application / x-tex">\circ< / annotation>< / semantics>< / com:Math> MTF Modulator<com:Sub>K< / com:Sub> 202, part of the digital side 510, which converts the (possibly FEC coded) information vector, <com:Math><semantics>α→K<annotation encoding="application / x-tex">\vec{\alpha}_K< / annotation>< / semantics>< / com:Math> 201, into an MTF (digital) vector, <com:Math><semantics>β→K<annotation encoding="application / x-tex">\vec{\beta}_K< / annotation>< / semantics>< / com:Math> 203, <com:Math><semantics>∘<annotation encoding="application / x-tex">\circ< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>TxK<annotation encoding="application / x-tex">Tx_K< / annotation>< / semantics>< / com:Math> 204, part of the digital side 510 and analog side 511, 513, which converts the MTF (digital) vector, <com:Math><semantics>β→K<annotation encoding="application / x-tex">\vec{\beta}_K< / annotation>< / semantics>< / com:Math> 203, into a transmitted MTF (analog) signal, <com:Math><semantics>xK(t)<annotation encoding="application / x-tex">x_K(t)< / annotation>< / semantics>< / com:Math> 205. The RU 321 comprises <com:Math><semantics>∘Nr<annotation encoding="application / x-tex">\circ N_r< / annotation>< / semantics>< / com:Math> Rx 308, ..., 208, part of the analog side 616, 618 and digital side 617, which converts the MTF (analog) signals, <com:Math><semantics>y1(t)<annotation encoding="application / x-tex">y_1(t)< / annotation>< / semantics>< / com:Math>, ..., <com:Math><semantics>yNr(t)<annotation encoding="application / x-tex">y_{N_r}(t)< / annotation>< / semantics>< / com:Math> 307, ..., 207 into <com:Math><semantics>Nr<annotation encoding="application / x-tex">N_r< / annotation>< / semantics>< / com:Math> received MTF (digital) vectors, <com:Math><semantics>r→1,...,r→Nr<annotation encoding="application / x-tex">\vec{r}_1, ..., \vec{r}_{N_r}< / annotation>< / semantics>< / com:Math> 309, ..., 209 respectively, and o MTF Detector<com:Sub>1< / com:Sub> 310, MTF Detector 110 part of the digital side 617, which detects the (possibly FEC coded) <com:Math><semantics>K<annotation encoding="application / x-tex">K< / annotation>< / semantics>< / com:Math> information vectors, <com:Math><semantics>α→1<annotation encoding="application / x-tex">\vec{\alpha}_1< / annotation>< / semantics>< / com:Math> 301, ..., <com:Math><semantics>α→K<annotation encoding="application / x-tex">\vec{\alpha}_K< / annotation>< / semantics>< / com:Math> 201, as <com:Math><semantics>α→̂1<annotation encoding="application / x-tex">\hat{\vec{\alpha}}_1< / annotation>< / semantics>< / com:Math> 311, ..., <com:Math><semantics>α→̂K<annotation encoding="application / x-tex">\hat{\vec{\alpha}}_K< / annotation>< / semantics>< / com:Math> 211, from the received MTF (digital) vectors <com:Math><semantics>r→1<annotation encoding="application / x-tex">\vec{r}_1< / annotation>< / semantics>< / com:Math> 109, ..., <com:Math><semantics>r→Nr<annotation encoding="application / x-tex">\vec{r}_{N_r}< / annotation>< / semantics>< / com:Math> 209, using CSIR 118, which is an estimate (possibly using a training sequence) of the state of the channel 406. • FIG. 5a is a schematic view of an embodiment of Tx 104 (204, 304) for converting a baseband- sampled (consisting of complex samples) or an IF-sampled (consisting of real samples) MTF (digital 510) discrete-time vector, <com:Math><semantics>β→<annotation encoding="application / x-tex">\vec{\beta}< / annotation>< / semantics>< / com:Math> 103 (203, 303), into an MTF (analog 511) continuous-time RF signal, <com:Math><semantics>x(t)<annotation encoding="application / x-tex">x(t)< / annotation>< / semantics>< / com:Math> 105 (205, 305), to be transmitted, Tx 104 (204, 304) comprising o a Pre-Channel Filter 500 to pre-filter the baseband-sampled or the IF-sampled MTF (digital 510) vector, <com:Math><semantics>β→<annotation encoding="application / x-tex">\vec{\beta}< / annotation>< / semantics>< / com:Math> 103 (203, 303), based on the knowledge of the statistics of the channel 106 (306, 406) represented by Channel State Information at Transmitter (CSIT) 215, at Tx 104 (204, 304), o a Digital-to-Analog (D-to-A) Converter 502 to convert the pre-channel baseband-sampled or IF-sampled filtered MTF (digital 510) vector 501 into an analog 511 baseband or IF MTF signal 503, o an Up-Converter 504 to convert the analog baseband or IF MTF signal 503 into an up- converted analog 511 RF MTF (passband) signal 505, o a Power Amplifier (PA) 506 to amplify the up-converted analog 511 RF signal 505 into an amplified up-converted analog RF MTF signal 507, and o an Analog Filter 508 to filter the amplified up-converted analog 511 RF MTF signal 507 into a filtered amplified up-converted analog 511 RF MTF signal <com:Math><semantics>x(t)<annotation encoding="application / x-tex">x(t)< / annotation>< / semantics>< / com:Math> 105 (205, 305), which complies with a pre-specified spectral mask 701. • FIG. 5b is a schematic view of an embodiment of Rx 108 (208, 308) for converting a received MTF (analog 616) continuous-time RF signal, <com:Math><semantics>y(t)<annotation encoding="application / x-tex">y(t)< / annotation>< / semantics>< / com:Math> 107 (207, 307), into a received baseband- sampled (consisting of complex samples) or IF-sampled (consisting of real samples), MTF (digital) discrete-time vector, <com:Math><semantics>r→<annotation encoding="application / x-tex">\vec{r}< / annotation>< / semantics>< / com:Math> 109 (209, 309), to be used for detecting the original (possibly FEC coded) information vector, <com:Math><semantics>α→<annotation encoding="application / x-tex">\vec{\alpha}< / annotation>< / semantics>< / com:Math> 101 (201, 301), Rx 108 (208, 308) comprising o an Analog Filter 607 to filter the received MTF (analog 616) RF signal, <com:Math><semantics>y(t)<annotation encoding="application / x-tex">y(t)< / annotation>< / semantics>< / com:Math> 107 (207, 307), into a filtered received analog RF MTF signal 608 (to filter and excise out-of-band noise and narrowband interferers), o a Low Noise Amplifier (LNA) 609 to amplify the filtered received analog 616 RF MTF signal 608, into an amplified filtered received analog 616 RF MTF signal 610, o a Down-Converter 611 to convert the amplified filtered received analog 616 RF MTF signal 610, into an amplified filtered received analog baseband or IF MTF signal 612, and o an Analog-to-Digital (A-to-D) Converter 613 to convert the amplified filtered received analog 616 baseband or IF MTF signal 612, into a baseband-sampled or IF-sampled digital 617 amplified filtered received MTF vector 614, o a Post-Channel Filter 615 to convert the baseband-sampled or IF-sampled digital 617 amplified filtered received MTF vector 614 into a post-channel filtered baseband-sampled or IF-sampled MTF vector <com:Math><semantics>r→<annotation encoding="application / x-tex">\vec{r}< / annotation>< / semantics>< / com:Math> 109 (209, 309) based on CSIR 115. • FIG. 6a is a schematic view of an alternative embodiment of Tx 104 for converting an RF- sampled, MTF (digital 510) discrete-time vector, <com:Math><semantics>β→<annotation encoding="application / x-tex">\vec{\beta}< / annotation>< / semantics>< / com:Math> 103 (203, 303), (consisting of real samples) into an (analog 513) continuous-time RF MTF signal, <com:Math><semantics>x(t)<annotation encoding="application / x-tex">x(t)< / annotation>< / semantics>< / com:Math> 105 (205, 305), to be transmitted, Tx 104 (204, 304) comprising o a Pre-Channel Filter 500 to pre-filter the RF sampled, pre-channel filtered, MTF (digital) vector, <com:Math><semantics>β→<annotation encoding="application / x-tex">\vec{\beta}< / annotation>< / semantics>< / com:Math> 103 (203, 303), based on CSIT 215 at Tx 104 (204, 304), o a D-to-A Converter 502 to convert the RF sampled, MTF (digital 510) vector 501, into an analog 513 RF MTF signal 503, o a PA 506 to amplify the analog 513 RF MTF signal 503, into an amplified analog 513 RF MTF signal 509, and o an Analog Filter 508 to filter the amplified analog 513 RF MTF signal 507, into a filtered amplified analog 513 RF MTF signal <com:Math><semantics>x(t)<annotation encoding="application / x-tex">x(t)< / annotation>< / semantics>< / com:Math> 105 (205, 305), which complies with a pre- specified spectral mask 701. • FIG. 6b is a schematic view of an alternative embodiment of Rx 108 (208, 308) for converting a received (analog 618) continuous-time RF MTF signal, <com:Math><semantics>y(t)<annotation encoding="application / x-tex">y(t)< / annotation>< / semantics>< / com:Math> 107 (207, 307), into a received RF MTF sampled, (digital 617) discrete-time MTF vector, <com:Math><semantics>r→<annotation encoding="application / x-tex">\vec{r}< / annotation>< / semantics>< / com:Math> 109 (209, 309), (consisting of real samples) to be used for detecting the original information vector, <com:Math><semantics>α→<annotation encoding="application / x-tex">\vec{\alpha}< / annotation>< / semantics>< / com:Math> 101 (201, 301), Rx 108 208, 308) comprising o an Analog Filter 607 to filter the received (analog 618) RF MTF signal, <com:Math><semantics>y(t)<annotation encoding="application / x-tex">y(t)< / annotation>< / semantics>< / com:Math> 107 (207, 307), into a filtered received analog 618 RF MTF signal 608 (to filter out-of-band noise and narrowband interferers), oan LNA 609 to amplify the filtered received analog 618 RF MTF signal 608, into an amplified filtered received analog 618 RF MTF signal 610, and o an A-to-D Converter 613 to convert the amplified filtered received analog 618 RF MTF signal 612, into an RF sampled digital 617 amplified filtered received MTF vector 614, o a Post-Channel Filter 215 to convert the RF sampled digital 617 amplified filtered received MTF vector 614 into a post-channel filtered RF sampled MTF vector <com:Math><semantics>r→<annotation encoding="application / x-tex">\vec{r}< / annotation>< / semantics>< / com:Math> 109 (209, 309) based on CSIR 115. • FIG. 7 shows a spectral mask, <com:Math><semantics>ℙWiFi(f)<annotation encoding="application / x-tex">\mathbb{P}_{WiFi}(f)< / annotation>< / semantics>< / com:Math>, 701 in dBr versus frequency, <com:Math><semantics>f−fc<annotation encoding="application / x-tex">f - f_c< / annotation>< / semantics>< / com:Math>, as an example (among many) of a mask, where <com:Math><semantics>fc<annotation encoding="application / x-tex">f_c< / annotation>< / semantics>< / com:Math> is the carrier frequency. The selected mask, <com:Math><semantics>ℙWiFi(f)<annotation encoding="application / x-tex">\mathbb{P}_{WiFi}(f)< / annotation>< / semantics>< / com:Math>, is designated for the IEEE802.11 (also known as WiFi) WLAN mask for a 20MHz band. The mask, <com:Math><semantics>ℙWiFi(f)<annotation encoding="application / x-tex">\mathbb{P}_{WiFi}(f)< / annotation>< / semantics>< / com:Math>, 701 contains three distinct (non-overlapping) spectral parts: 1. the Occupied Band 705, 2. the Out-of-Band Emissions (OOBE) Band 706, and 3. the Far Out Spurious Emmissions (FOSE) Band 707. Many other mask constraints are available depending on the standard, frequency band and jurisdiction. While different, all masks must generally contain three distinct spectral parts: Occupied Band 705, OOBE Band 706 and FOSE Band 707, similar to the ones described above for <com:Math><semantics>ℙWiFi(f)<annotation encoding="application / x-tex">\mathbb{P}_{WiFi}(f)< / annotation>< / semantics>< / com:Math> 701. • FIG. 8 shows the PSD, <com:Math><semantics>Sx(t)(f)<annotation encoding="application / x-tex">S_{x(t)}(f)< / annotation>< / semantics>< / com:Math> 802, in dBr versus frequency, <com:Math><semantics>f−fc<annotation encoding="application / x-tex">f - f_c< / annotation>< / semantics>< / com:Math>, of the MTF signal, <com:Math><semantics>x(t)<annotation encoding="application / x-tex">x(t)< / annotation>< / semantics>< / com:Math> 105 (205, 305), where <com:Math><semantics>fc<annotation encoding="application / x-tex">f_c< / annotation>< / semantics>< / com:Math> is the carrier frequency. The MTF signal, <com:Math><semantics>x(t)<annotation encoding="application / x-tex">x(t)< / annotation>< / semantics>< / com:Math> 105 (205, 305), is designed in an attempt to match its PSD, <com:Math><semantics>Sx(t)(f)<annotation encoding="application / x-tex">S_{x(t)}(f)< / annotation>< / semantics>< / com:Math> 802, with the spectral mask, <com:Math><semantics>ℙWiFi(f)<annotation encoding="application / x-tex">\mathbb{P}_{WiFi}(f)< / annotation>< / semantics>< / com:Math> 701, in FIG. 7 as much as possible, however, without exceeding it. The PSD, <com:Math><semantics>Sx(t)(f)<annotation encoding="application / x-tex">S_{x(t)}(f)< / annotation>< / semantics>< / com:Math> 802, contains two distinct spectral parts: <com:Math><semantics>∘S℘k,ℵk𝒫(t)(f)<annotation encoding="application / x-tex">\circ S_{\wp_{k,\aleph_k}^{\mathcal{P}}(t)}(f)< / annotation>< / semantics>< / com:Math> 808, which is the PSD of <com:Math><semantics>℘→k,ℵk𝒫(f)<annotation encoding="application / x-tex">\vec{\wp}_{k,\aleph_k}^{\mathcal{P}}(f)< / annotation>< / semantics>< / com:Math>, to be matched with both the Occupied Band 705 and the OOBE Band 706 of the spectral mask, <com:Math><semantics>ℙWiFi(f)<annotation encoding="application / x-tex">\mathbb{P}_{WiFi}(f)< / annotation>< / semantics>< / com:Math> 701, as much as possible, but without exceeding it. <com:Math><semantics>∘𝒮℘k,0𝒫(t)(f)<annotation encoding="application / x-tex">\circ \mathcal{S}_{\wp_{k,0}^{\mathcal{P}}(t)}(f)< / annotation>< / semantics>< / com:Math> 809, which is the PSD of <com:Math><semantics>p→k,0𝒫<annotation encoding="application / x-tex">\vec{p}_{k,0}^{\mathcal{P}}< / annotation>< / semantics>< / com:Math>, to be matched with the FOSE Band 707 of the spectral mask, <com:Math><semantics>ℙWiFi(f)<annotation encoding="application / x-tex">\mathbb{P}_{WiFi}(f)< / annotation>< / semantics>< / com:Math> 701, as much as possible, but without exceeding it. • FIG. 9 displays <com:Math><semantics>S℘k,ℵk𝒫(t)(f)<annotation encoding="application / x-tex">S_{\wp_{k,\aleph_{k}}^{\mathcal{P}}(t)}(f)< / annotation>< / semantics>< / com:Math> 808 for <com:Math><semantics>MTF3<annotation encoding="application / x-tex">MTF_{3}< / annotation>< / semantics>< / com:Math> with a null at the restricted bands as well as across <com:Math><semantics>|f|≤2GHz<annotation encoding="application / x-tex">|f| \le 2GHz< / annotation>< / semantics>< / com:Math>, when <com:Math><semantics>℘→k,ℵk𝒫<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,\aleph_k}^{\mathcal{P}}< / annotation>< / semantics>< / com:Math> is selected as <com:Math><semantics>℘→REC,1𝒫<annotation encoding="application / x-tex">\overrightarrow{\wp}_{REC,1}^{\mathcal{P}}< / annotation>< / semantics>< / com:Math> with a PSD <com:Math><semantics>𝒮℘REC,1𝒫(t)(f)<annotation encoding="application / x-tex">\mathcal{S}_{\wp_{REC,1}^{\mathcal{P}}(t)}(f)< / annotation>< / semantics>< / com:Math> 908 while <com:Math><semantics>℘→k,0𝒫<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,0}^{\mathcal{P}}< / annotation>< / semantics>< / com:Math> is selected as defined in (17) with a PSD <com:Math><semantics>𝒮℘k0𝒫(t)(f)<annotation encoding="application / x-tex">\mathcal{S}_{\wp_{k_0}^{\mathcal{P}}(t)}(f)< / annotation>< / semantics>< / com:Math> 809 pre-distorted according to (20). From FIG. 9, one can deduce that it is possible to increase <com:Math><semantics>𝒞MTF<annotation encoding="application / x-tex">\mathcal{C}_{MTF}< / annotation>< / semantics>< / com:Math> arbitrarily by increasing <com:Math><semantics>fs3<annotation encoding="application / x-tex">f_{s_3}< / annotation>< / semantics>< / com:Math> regardless of the value of <com:Math><semantics>fc<annotation encoding="application / x-tex">f_c< / annotation>< / semantics>< / com:Math> where <com:Math><semantics>fs3<annotation encoding="application / x-tex">f_{s_3}< / annotation>< / semantics>< / com:Math> is the sampling frequency of <com:Math><semantics>MTF3<annotation encoding="application / x-tex">MTF_3< / annotation>< / semantics>< / com:Math>. • FIG. 10a compares <com:Math><semantics>𝒞TL / 𝒲m<annotation encoding="application / x-tex">\mathcal{C}_{TL} / \mathcal{W}_m< / annotation>< / semantics>< / com:Math> in (5) (shown with "." markers) with <com:Math><semantics>𝒞BL / 𝒲m<annotation encoding="application / x-tex">\mathcal{C}_{BL} / \mathcal{W}_m< / annotation>< / semantics>< / com:Math> in (6) (shown with "*" markers) when <com:Math><semantics>r=rBL=1<annotation encoding="application / x-tex">r = r_{BL} = 1< / annotation>< / semantics>< / com:Math>, <com:Math><semantics>d=K=1<annotation encoding="application / x-tex">d = K = 1< / annotation>< / semantics>< / com:Math> and the <com:Math><semantics>kth<annotation encoding="application / x-tex">k^{th}< / annotation>< / semantics>< / com:Math> column, <com:Math><semantics>{hCh}k<annotation encoding="application / x-tex">\{h_{Ch}\}_k< / annotation>< / semantics>< / com:Math> of <com:Math><semantics>hCh<annotation encoding="application / x-tex">h_{Ch}< / annotation>< / semantics>< / com:Math> corresponds to a rectangular pulse <com:Math><semantics>∀k<annotation encoding="application / x-tex">\forall k< / annotation>< / semantics>< / com:Math>. In FIG. 10a, <com:Math><semantics>𝒞TL / 𝒲m<annotation encoding="application / x-tex">\mathcal{C}_{TL} / \mathcal{W}_m< / annotation>< / semantics>< / com:Math> in (5) is illustrated as a number of curves, each curve corresponding to a value of <com:Math><semantics>Nmin<annotation encoding="application / x-tex">N_{min}< / annotation>< / semantics>< / com:Math>. The selected values are <com:Math><semantics>Nmin=1,2,4,...,1024<annotation encoding="application / x-tex">N_{min} = 1, 2, 4, ..., 1024< / annotation>< / semantics>< / com:Math>, with <com:Math><semantics>Nmin=1<annotation encoding="application / x-tex">N_{min} = 1< / annotation>< / semantics>< / com:Math> coinciding with <com:Math><semantics>CBL / Wm<annotation encoding="application / x-tex">C_{BL} / W_m< / annotation>< / semantics>< / com:Math> in (6). <com:Math><semantics>CBL / Wm<annotation encoding="application / x-tex">C_{BL} / W_m< / annotation>< / semantics>< / com:Math> in (6) contains a "Low" SNR region 1004 and a "High" SNR region 1005. Similarly, <com:Math><semantics>CTL / Wm<annotation encoding="application / x-tex">C_{TL} / W_m< / annotation>< / semantics>< / com:Math> in (5) contains a "Low" SNR region 1001 and a "High" SNR region 1003. Unlike <com:Math><semantics>CBL / Wm<annotation encoding="application / x-tex">C_{BL} / W_m< / annotation>< / semantics>< / com:Math> in (6), <com:Math><semantics>CTL / Wm<annotation encoding="application / x-tex">C_{TL} / W_m< / annotation>< / semantics>< / com:Math> in (5) also contains a medium SNR region 1002, denoted as "Med. SNR" in FIG. 10a where doubling <com:Math><semantics>𝒞TL / 𝒲m<annotation encoding="application / x-tex">\mathcal{C}_{TL} / \mathcal{W}_m< / annotation>< / semantics>< / com:Math> in (5) requires a fixed multiple increase in <com:Math><semantics>ΛkrNmin𝒜̃K𝔭ℕo𝒲m<annotation encoding="application / x-tex">\frac{\Lambda_k r}{N_{min}} \frac{\tilde{\mathcal{A}} K \mathfrak{p}}{\mathbb{N}_o \mathcal{W}_m}< / annotation>< / semantics>< / com:Math> since its contribution towards <com:Math><semantics>𝒞TL<annotation encoding="application / x-tex">\mathcal{C}_{TL}< / annotation>< / semantics>< / com:Math> is mostly linear. • FIG. 10b shows that based on (10), <com:Math><semantics>𝒞MTF<annotation encoding="application / x-tex">\mathcal{C}_{MTF}< / annotation>< / semantics>< / com:Math> consists of two medium SNR regions 1006, 1007. As N increases, doubling <com:Math><semantics>CMTF<annotation encoding="application / x-tex">C_{MTF}< / annotation>< / semantics>< / com:Math> initially requires increasing the SNR by a fixed multiple of <com:Math><semantics>22ℵ<annotation encoding="application / x-tex">2^{2\aleph}< / annotation>< / semantics>< / com:Math>, which is the first medium SNR region 1006, referred to as "1st Med. SNR" in FIG. 10b. Then, later on it requires increasing the SNR by a fixed multiple of 2, which is the second medium SNR region 1007, referred to as "2<com:Sup>nd< / com:Sup> Med. SNR". FIG. 10b compares <com:Math><semantics>𝒞MTF / 𝒲m<annotation encoding="application / x-tex">\mathcal{C}_{MTF} / \mathcal{W}_m< / annotation>< / semantics>< / com:Math> based on (10) with <com:Math><semantics>CBL / Wm<annotation encoding="application / x-tex">C_{BL} / W_m< / annotation>< / semantics>< / com:Math> in (6), where <com:Math><semantics>r=rBL=1<annotation encoding="application / x-tex">r = r_{BL} = 1< / annotation>< / semantics>< / com:Math>, <com:Math><semantics>ℵ=1<annotation encoding="application / x-tex">\aleph = 1< / annotation>< / semantics>< / com:Math>, <com:Math><semantics>d=K=1<annotation encoding="application / x-tex">d = K = 1< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>℘→k,ℵ<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,\aleph}< / annotation>< / semantics>< / com:Math> in (8) is a rectangular pulse with <com:Math><semantics>℘→k,0<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,0}< / annotation>< / semantics>< / com:Math> in (8) <com:Math><semantics>30dBr<annotation encoding="application / x-tex">30dBr< / annotation>< / semantics>< / com:Math> below <com:Math><semantics>℘→k,N<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,N}< / annotation>< / semantics>< / com:Math>. In FIG. 10b, <com:Math><semantics>𝒞MTF / 𝒲m<annotation encoding="application / x-tex">\mathcal{C}_{MTF} / \mathcal{W}_m< / annotation>< / semantics>< / com:Math> without Constraint 3 is illustrated as a number of curves (with "." markers), each curve corresponding to a value of <com:Math><semantics>Nmin<annotation encoding="application / x-tex">N_{min}< / annotation>< / semantics>< / com:Math>. <com:Math><semantics>CMTF / Wm<annotation encoding="application / x-tex">C_{MTF} / W_m< / annotation>< / semantics>< / com:Math> with Constraint 3 is illustrated as a number of points (with "square" markers), each point corresponding to a value of <com:Math><semantics>Nmin<annotation encoding="application / x-tex">N_{min}< / annotation>< / semantics>< / com:Math>. FIG. 10b shows that indeed Constraint 3 maximizes <com:Math><semantics>CMTF / Wm<annotation encoding="application / x-tex">C_{MTF} / W_m< / annotation>< / semantics>< / com:Math> in both medium SNR regions 1006, 1007. • FIG. 11a shows an embodiment of <com:Math><semantics>h<annotation encoding="application / x-tex">h< / annotation>< / semantics>< / com:Math> 1101 as a block toeplitz matrix. The sub-matrix <com:Math><semantics>h→≜<annotation encoding="application / x-tex">\vec{h} \triangleq< / annotation>< / semantics>< / com:Math> <com:Math><semantics>[𝒉Basic𝟎A(I=1)×K]∈ℂM×K<annotation encoding="application / x-tex">\begin{bmatrix} \mathbf{h}_{Basic} \\ \mathbf{0}_{A(I=1)\times K} \end{bmatrix} \in \mathbb{C}^{M\times K}< / annotation>< / semantics>< / com:Math> is used as a building block of <com:Math><semantics>𝒉<annotation encoding="application / x-tex">\mathbf{h}< / annotation>< / semantics>< / com:Math> 1101 where <com:Math><semantics>𝒉Basic∈ℂN×K<annotation encoding="application / x-tex">\mathbf{h}_{Basic} \in \mathbb{C}^{N\times K}< / annotation>< / semantics>< / com:Math> in FIG. 11a comprises <com:Math><semantics>Nd<annotation encoding="application / x-tex">N_d< / annotation>< / semantics>< / com:Math> sub-blocks with the first <com:Math><semantics>|Nd|<annotation encoding="application / x-tex">\left|\frac{N}{d}\right|< / annotation>< / semantics>< / com:Math> sub-blocks consisting of <com:Math><semantics>d<annotation encoding="application / x-tex">d< / annotation>< / semantics>< / com:Math> rows each, while the last sub-block consisting of <com:Math><semantics>d′<annotation encoding="application / x-tex">d'< / annotation>< / semantics>< / com:Math> rows, i.e. <com:Math><semantics>𝒉Basic=[𝒉Basic,1⋯𝒉Basic,Nd]T<annotation encoding="application / x-tex">\mathbf{h}_{Basic} = [\mathbf{h}_{Basic,1} \cdots \mathbf{h}_{Basic,N_d}]^T< / annotation>< / semantics>< / com:Math> with <com:Math><semantics>𝒉Basic,1∈<annotation encoding="application / x-tex">\mathbf{h}_{Basic,1} \in< / annotation>< / semantics>< / com:Math> <com:Math><semantics>ℂd×K<annotation encoding="application / x-tex">\mathbb{C}^{d\times K}< / annotation>< / semantics>< / com:Math> 1102 and <com:Math><semantics>hBasic,Nd∈ℂd′×K<annotation encoding="application / x-tex">h_{Basic,N_d} \in \mathbb{C}^{d'\times K}< / annotation>< / semantics>< / com:Math> 1103, where <com:Math><semantics>d′≜Nmodulo(d)<annotation encoding="application / x-tex">d' \triangleq N \ modulo(d)< / annotation>< / semantics>< / com:Math>, <com:Math><semantics>Nd≜|Nd|<annotation encoding="application / x-tex">N_d \triangleq \left|\frac{N}{d}\right|< / annotation>< / semantics>< / com:Math> is the ceiling function, <com:Math><semantics>|Nd|<annotation encoding="application / x-tex">\left|\frac{N}{d}\right|< / annotation>< / semantics>< / com:Math> is the floor function and [.]<com:Sup>T< / com:Sup> denotes a transpose operation. • FIG. 11b shows an embodiment of <com:Math><semantics>hMTF<annotation encoding="application / x-tex">h_{MTF}< / annotation>< / semantics>< / com:Math> 1104 as a block toeplitz matrix, where the K TUs 316, 318 are replaced by one equivalent TU 314 with one equivalent <com:Math><semantics>hMTF<annotation encoding="application / x-tex">h_{MTF}< / annotation>< / semantics>< / com:Math> 1104. The sub- matrix <com:Math><semantics>h→MTF≜[hB_MTF𝟎A(I=1)×K]∈ℂM×K<annotation encoding="application / x-tex">\vec{h}_{MTF} \triangleq \begin{bmatrix} h_{B\_MTF} \\ \mathbf{0}_{A(I=1)\times K} \end{bmatrix} \in \mathbb{C}^{M\times K}< / annotation>< / semantics>< / com:Math> is used as a building block of <com:Math><semantics>hMTF<annotation encoding="application / x-tex">h_{MTF}< / annotation>< / semantics>< / com:Math> 1104 where <com:Math><semantics>hB_MTF<annotation encoding="application / x-tex">h_{B\_MTF}< / annotation>< / semantics>< / com:Math> in FIG. 11b comprises <com:Math><semantics>Nd<annotation encoding="application / x-tex">N_d< / annotation>< / semantics>< / com:Math> sub-blocks with the first <com:Math><semantics>|Nd|<annotation encoding="application / x-tex">\left|\frac{N}{d}\right|< / annotation>< / semantics>< / com:Math> sub-blocks consisting of d rows each, while the last sub-block consisting of <com:Math><semantics>d′<annotation encoding="application / x-tex">d'< / annotation>< / semantics>< / com:Math> rows, i.e. <com:Math><semantics>𝒉B_MTF=[𝒉B_MTF,1⋯𝒉B_MTF,Nd]T<annotation encoding="application / x-tex">\mathbf{h}_{B\_MTF} = \left[\mathbf{h}_{B\_MTF,1} \cdots \mathbf{h}_{B\_MTF,N_d}\right]^T< / annotation>< / semantics>< / com:Math> with <com:Math><semantics>hBMTF,1∈ℂd×K<annotation encoding="application / x-tex">h_{B\ MTF,1} \in \mathbb{C}^{d \times K}< / annotation>< / semantics>< / com:Math>1105 and <com:Math><semantics>hBMTF,Nd∈ℂd′×K<annotation encoding="application / x-tex">h_{B\ MTF,N_d} \in \mathbb{C}^{d' \times K}< / annotation>< / semantics>< / com:Math>1106. • FIG. 12a displays an embodiment for implementing equations (8) and (9). Equation (9) <com:Math><semantics>∘<annotation encoding="application / x-tex">\circ< / annotation>< / semantics>< / com:Math> convolves <com:Math><semantics>h1,k<annotation encoding="application / x-tex">h_{1,k}< / annotation>< / semantics>< / com:Math> 1201 with <com:Math><semantics>g→1,k<annotation encoding="application / x-tex">\vec{g}_{1,k}< / annotation>< / semantics>< / com:Math> 1203 using a circular convolution, denoted by "* 1202, and o convolves <com:Math><semantics>h→ℵk,k<annotation encoding="application / x-tex">\vec{h}_{\aleph_k,k}< / annotation>< / semantics>< / com:Math> 1204 with <com:Math><semantics>g→ℵk,k<annotation encoding="application / x-tex">\vec{g}_{\aleph_k,k}< / annotation>< / semantics>< / com:Math> 1206 using a circular convolution, denoted by “<com:Math><semantics>⊛<annotation encoding="application / x-tex">\circledast< / annotation>< / semantics>< / com:Math>” 1205, othen, it performs <com:Math><semantics>ℵk−1<annotation encoding="application / x-tex">\aleph_k - 1< / annotation>< / semantics>< / com:Math> linear convolution operations, <com:Math><semantics>(h→1,k⊛g→1,k)<annotation encoding="application / x-tex">(\vec{h}_{1,k} \circledast \vec{g}_{1,k})< / annotation>< / semantics>< / com:Math> 1215 * ... * <com:Math><semantics>(h→ℵk,k⊗g→ℵk,k)<annotation encoding="application / x-tex">(\vec{h}_{\aleph_k,k} \otimes \vec{g}_{\aleph_k,k})< / annotation>< / semantics>< / com:Math> 1216, using a linear convolution operator, denoted by "*" 1207, in order to produce <com:Math><semantics>℘→k,ℵν<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,\aleph_{\nu}}< / annotation>< / semantics>< / com:Math> 1208. Equation (8) adds <com:Math><semantics>℘→k,ℵk<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,\aleph_k}< / annotation>< / semantics>< / com:Math> 1208 that is obtained from equation (9) with <com:Math><semantics>℘→k,0<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,0}< / annotation>< / semantics>< / com:Math> 1209 using an adder denoted by "+" 1210 in order to produce <com:Math><semantics>{hB_MTF}k<annotation encoding="application / x-tex">\{h_{B\_MTF}\}_{k}< / annotation>< / semantics>< / com:Math> 1308. • FIG. 12b displays a preferred embodiment for implementing equations (8) and (9) where all <com:Math><semantics>h→l,k∀l1201,1204<annotation encoding="application / x-tex">\vec{h}_{l,k} \ \forall l \ 1201, \ 1204< / annotation>< / semantics>< / com:Math> in FIG. 12a are selected as <com:Math><semantics>h→REC,l,k1212,1213<annotation encoding="application / x-tex">\vec{h}_{REC,l,k} \ 1212, \ 1213< / annotation>< / semantics>< / com:Math>. In this case, <com:Math><semantics>℘→k,ℵk1208<annotation encoding="application / x-tex">\vec{\wp}_{k,\aleph_k} \ 1208< / annotation>< / semantics>< / com:Math> in (8) is denoted as <com:Math><semantics>℘→REC,ℵk<annotation encoding="application / x-tex">\overrightarrow{\wp}_{REC,\aleph_k}< / annotation>< / semantics>< / com:Math> 1214. • FIG. 13 displays a preferred embodiment 1309 for implementing MTF Design Steps I-III, which designs <com:Math><semantics>{hB_MTF}k<annotation encoding="application / x-tex">\{h_{B\_MTF}\}_{k}< / annotation>< / semantics>< / com:Math> 1308 based on Theorem III with the goal of achieving a desired channel capacity, <com:Math><semantics>Cd<annotation encoding="application / x-tex">C_d< / annotation>< / semantics>< / com:Math>, for a communications channel with <com:Math><semantics>(NTs)min<annotation encoding="application / x-tex">(NT_s)_{min}< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>Ts,max<annotation encoding="application / x-tex">T_{s,max}< / annotation>< / semantics>< / com:Math> 1301 defined: <com:Math><semantics>∘<annotation encoding="application / x-tex">\circ< / annotation>< / semantics>< / com:Math> MTF Design Step I 1302: This step is referred to as the selection step 1302. It accepts <com:Math><semantics>𝒞d<annotation encoding="application / x-tex">\mathcal{C}_d< / annotation>< / semantics>< / com:Math>, <com:Math><semantics>Nmin<annotation encoding="application / x-tex">N_{min}< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>Ts,max<annotation encoding="application / x-tex">T_{s,max}< / annotation>< / semantics>< / com:Math> 1301, and generates the N selected FAT DOF 1303. • MTF Design Step II 1304: This step is referred to as the enhancement step 1304. It accepts the N selected FAT DOF 1303, and generates the N selected and enhanced FAT DOF 1305. o MTF Design Step III 1306: This step is referred to as the randomization step 1306. It accepts the N selected and enhanced FAT DOF 1305, and generates the N selected, enhanced and randomized FAT DOF 1307. The N selected, enhanced and randomized FAT DOF 1307 are used to form the N elements of the <com:Math><semantics>kth<annotation encoding="application / x-tex">k^{th}< / annotation>< / semantics>< / com:Math> column, <com:Math><semantics>{hB_MTF}l<annotation encoding="application / x-tex">\{h_{B\_MTF}\}_{l}< / annotation>< / semantics>< / com:Math> 1308, of <com:Math><semantics>hB_MTF<annotation encoding="application / x-tex">h_{B\_MTF}< / annotation>< / semantics>< / com:Math> using an inverse transform 1310. 5 DETAILED DESCRIPTION OF INVENTION 5.1 TL SYSTEMS An information vector, <com:Math><semantics>α→∈ℂQ×1<annotation encoding="application / x-tex">\vec{\alpha} \in \mathbb{C}^{Q \times 1}< / annotation>< / semantics>< / com:Math> 101, 201, 301, consisting of <com:Math><semantics>Q<annotation encoding="application / x-tex">Q< / annotation>< / semantics>< / com:Math> (possibly FEC coded) information symbols, can be transmitted by one or several active transmitters, Tx 104, 204, 304, across a communications channel 106, 306, 406 by converting <com:Math><semantics>α→<annotation encoding="application / x-tex">\vec{\alpha}< / annotation>< / semantics>< / com:Math> 101, 201, 301 into a vector, <com:Math><semantics>β→<annotation encoding="application / x-tex">\vec{\beta}< / annotation>< / semantics>< / com:Math> 103, 203, 303, defined as [Image disponible dans le document PDF, Image available in the PDF document] (1) using a matrix, <com:Math><semantics>h∈ℂM×Q<annotation encoding="application / x-tex">h \in \mathbb{C}^{M \times Q}< / annotation>< / semantics>< / com:Math> 1101, where <com:Math><semantics>β→∈ℂM×1<annotation encoding="application / x-tex">\vec{\beta} \in \mathbb{C}^{M \times 1}< / annotation>< / semantics>< / com:Math> 103, 203, 303 consists of samples, each of duration <com:Math><semantics>Ts<annotation encoding="application / x-tex">T_s< / annotation>< / semantics>< / com:Math> for a total duration for <com:Math><semantics>β→<annotation encoding="application / x-tex">\vec{\beta}< / annotation>< / semantics>< / com:Math> of <com:Math><semantics>MTs<annotation encoding="application / x-tex">MT_s< / annotation>< / semantics>< / com:Math>. In this disclosure, we select h 1101, to be block toeplitz, i.e. h 1101 is defined as [Image disponible dans le document PDF, Image available in the PDF document] (2) where <com:Math><semantics>toepd{h→}<annotation encoding="application / x-tex">toep_d\{\vec{h}\}< / annotation>< / semantics>< / com:Math> is an operator, which forms <com:Math><semantics>h<annotation encoding="application / x-tex">h< / annotation>< / semantics>< / com:Math> 1101 by repeatedly replicating the sub-matrix <com:Math><semantics>h→<annotation encoding="application / x-tex">\vec{h}< / annotation>< / semantics>< / com:Math> to the right <com:Math><semantics>L−1<annotation encoding="application / x-tex">L-1< / annotation>< / semantics>< / com:Math> times, while cyclically shifting <com:Math><semantics>h→<annotation encoding="application / x-tex">\vec{h}< / annotation>< / semantics>< / com:Math> down by <com:Math><semantics>d<annotation encoding="application / x-tex">d< / annotation>< / semantics>< / com:Math> rows for every single replica to the right, with <com:Math><semantics>L≜[Q / K]<annotation encoding="application / x-tex">L \triangleq [Q / K]< / annotation>< / semantics>< / com:Math>, defined as the ceiling of <com:Math><semantics>Q / K<annotation encoding="application / x-tex">Q / K< / annotation>< / semantics>< / com:Math>. <com:Math><semantics>h→∈ℂM×K<annotation encoding="application / x-tex">\vec{h} \in \mathbb{C}^{M \times K}< / annotation>< / semantics>< / com:Math> is defined as <com:Math><semantics>h→≜<annotation encoding="application / x-tex">\vec{h} \triangleq< / annotation>< / semantics>< / com:Math> <com:Math><semantics>∣𝒏Basic𝟎d(L−1)×K∣<annotation encoding="application / x-tex">\begin{vmatrix} \mathbf{n}_{Basic} \\ \mathbf{0}_{d(L-1)\times K} \end{vmatrix}< / annotation>< / semantics>< / com:Math> where <com:Math><semantics>𝒉Basic∈ℂN×K<annotation encoding="application / x-tex">\mathbf{h}_{Basic} \in \mathbb{C}^{N\times K}< / annotation>< / semantics>< / com:Math> is referred to as the basic building block, while <com:Math><semantics>𝟎d(L−1)×K<annotation encoding="application / x-tex">\mathbf{0}_{d(L-1)\times K}< / annotation>< / semantics>< / com:Math> is the all zero <com:Math><semantics>d(L−1)×K<annotation encoding="application / x-tex">d(L-1) \times K< / annotation>< / semantics>< / com:Math> matrix, with <com:Math><semantics>d≤N<annotation encoding="application / x-tex">d \leq N< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>M≜N+d(L−1)<annotation encoding="application / x-tex">M \triangleq N + d(L-1)< / annotation>< / semantics>< / com:Math>. Interpretation of h 1101: Since each column of h 1101 is responsible for transporting one information symbol in <com:Math><semantics>α→<annotation encoding="application / x-tex">\vec{\alpha}< / annotation>< / semantics>< / com:Math> 101, 201, 301, therefore, <com:Math><semantics>β→<annotation encoding="application / x-tex">\vec{\beta}< / annotation>< / semantics>< / com:Math> 103, 203, 303 in (1) can model the output of a <com:Math><semantics>K−user<annotation encoding="application / x-tex">K-user< / annotation>< / semantics>< / com:Math> TL system with spreading gain <com:Math><semantics>N<∞<annotation encoding="application / x-tex">N < \infty< / annotation>< / semantics>< / com:Math> with a number, <com:Math><semantics>Kb<annotation encoding="application / x-tex">K_b< / annotation>< / semantics>< / com:Math>, of desired transmitters (Txs) intended for a receiver, Rx 108, 308, and a number, <com:Math><semantics>Ki<annotation encoding="application / x-tex">K_i< / annotation>< / semantics>< / com:Math>, of interfering Txs 204, s.t. <com:Math><semantics>K=100<annotation encoding="application / x-tex">K = 100< / annotation>< / semantics>< / com:Math> <com:Math><semantics>Kb+Ki<annotation encoding="application / x-tex">K_{b} + K_{i}< / annotation>< / semantics>< / com:Math>. The <com:Math><semantics>kth<annotation encoding="application / x-tex">k^{th}< / annotation>< / semantics>< / com:Math> active Tx 104, 204, 304 transmits a vector <com:Math><semantics>β→k<annotation encoding="application / x-tex">\vec{\beta}_{k}< / annotation>< / semantics>< / com:Math>, which transports the set of L symbols, <com:Math><semantics>{αk,αk+K,…,αk+(L−1)K}<annotation encoding="application / x-tex">\{\alpha_k, \alpha_{k+K}, \dots, \alpha_{k+(L-1)K}\}< / annotation>< / semantics>< / com:Math>, after converting <com:Math><semantics>β→k<annotation encoding="application / x-tex">\vec{\beta}_k< / annotation>< / semantics>< / com:Math> into a continuous-time signal, <com:Math><semantics>xk(t)<annotation encoding="application / x-tex">x_k(t)< / annotation>< / semantics>< / com:Math>, of finite duration <com:Math><semantics>MTs<annotation encoding="application / x-tex">MT_s< / annotation>< / semantics>< / com:Math>, with <com:Math><semantics>Ts<annotation encoding="application / x-tex">T_s< / annotation>< / semantics>< / com:Math> the duration of one sample in <com:Math><semantics>β→<annotation encoding="application / x-tex">\vec{\beta}< / annotation>< / semantics>< / com:Math> 103, 203, 303. Theorem I assumes that: 1. The <com:Math><semantics>kth<annotation encoding="application / x-tex">k^{th}< / annotation>< / semantics>< / com:Math> Tx 104, 204, 304 transmits <com:Math><semantics>xk(t)<annotation encoding="application / x-tex">x_k(t)< / annotation>< / semantics>< / com:Math> 105, 205, 305 subject to Constraint 1: [Image disponible dans le document PDF, Image available in the PDF document] where p is the average allocatable transmit power at any Tx 104, 204, 304 and <com:Math><semantics>Sxk(t)(f)<annotation encoding="application / x-tex">S_{x_k(t)}(f)< / annotation>< / semantics>< / com:Math> 802 is the Power Spectral Density (PSD) of <com:Math><semantics>xk(t)<annotation encoding="application / x-tex">x_k(t)< / annotation>< / semantics>< / com:Math> 105, 205, 305. 2. The <com:Math><semantics>kth<annotation encoding="application / x-tex">k^{th}< / annotation>< / semantics>< / com:Math> Tx 104, 204, 304 transmits <com:Math><semantics>xk(t)<annotation encoding="application / x-tex">x_k(t)< / annotation>< / semantics>< / com:Math> 105, 205, 305 using a single antenna across a communication channel where it is received at Rx 108, 208, 308 using a single antenna. The received signal, <com:Math><semantics>y(t)<annotation encoding="application / x-tex">y(t)< / annotation>< / semantics>< / com:Math>, is then sampled at Rx 108, 208, 308 at a sampling frequency, <com:Math><semantics>fs≜1 / Ts<annotation encoding="application / x-tex">f_s \triangleq 1 / T_s< / annotation>< / semantics>< / com:Math>, to form a discrete-time signal <com:Math><semantics>r→∈ℂM×1<annotation encoding="application / x-tex">\vec{r} \in \mathbb{C}^{M \times 1}< / annotation>< / semantics>< / com:Math> defined as [Image disponible dans le document PDF, Image available in the PDF document] (3) where <com:Math><semantics>hCh∈ℂM×Q<annotation encoding="application / x-tex">h_{Ch} \in \mathbb{C}^{M \times Q}< / annotation>< / semantics>< / com:Math> corresponds to <com:Math><semantics>h<annotation encoding="application / x-tex">h< / annotation>< / semantics>< / com:Math> 1101 after including the effects of the channel 106, 306, 406 such as replacing M by a number <com:Math><semantics>ℳ≥M<annotation encoding="application / x-tex">\mathcal{M} \geq M< / annotation>< / semantics>< / com:Math>, and <com:Math><semantics>w→∈ℂM×1<annotation encoding="application / x-tex">\vec{w} \in \mathbb{C}^{M \times 1}< / annotation>< / semantics>< / com:Math> models the WGN. We refer to the combination of the TL system and channel as a TL channel. Theorem I: The capacity, <com:Math><semantics>CTL<annotation encoding="application / x-tex">C_{TL}< / annotation>< / semantics>< / com:Math>, of the TL channel corresponding to <com:Math><semantics>hCh<annotation encoding="application / x-tex">h_{Ch}< / annotation>< / semantics>< / com:Math> in (3), subject to Constraint 1, is [Image disponible dans le document PDF, Image available in the PDF document] (4) where <com:Math><semantics>ℕo / 2<annotation encoding="application / x-tex">\mathbb{N}_o / 2< / annotation>< / semantics>< / com:Math> is the two-sided PSD of the WGN, <com:Math><semantics>𝒜‾<annotation encoding="application / x-tex">\bar{\mathcal{A}}< / annotation>< / semantics>< / com:Math> is the average attenuation in power across the channel, and <com:Math><semantics>Λk<annotation encoding="application / x-tex">\Lambda_k< / annotation>< / semantics>< / com:Math> is the <com:Math><semantics>kth<annotation encoding="application / x-tex">k^{th}< / annotation>< / semantics>< / com:Math> squared singular value of a normalized <com:Math><semantics>hCh<annotation encoding="application / x-tex">h_{Ch}< / annotation>< / semantics>< / com:Math> s.t. its <com:Math><semantics>kth<annotation encoding="application / x-tex">k^{th}< / annotation>< / semantics>< / com:Math> column, <com:Math><semantics>{𝒉Ch}k<annotation encoding="application / x-tex">\{\boldsymbol{h}_{Ch}\}_k< / annotation>< / semantics>< / com:Math>, has on average an L2-norm, which equals <com:Math><semantics>d∀k<annotation encoding="application / x-tex">d \forall k< / annotation>< / semantics>< / com:Math>. Importance of Theorem I: <com:Math><semantics>CTL<annotation encoding="application / x-tex">C_{TL}< / annotation>< / semantics>< / com:Math> in (4) consists of several regions, which depend on the average received TL SNR, <com:Math><semantics>𝒜K𝔭Nαfα<annotation encoding="application / x-tex">\frac{\mathcal{A}K\mathfrak{p}}{N_{\alpha}f_{\alpha}}< / annotation>< / semantics>< / com:Math>. Similar to the capacity, <com:Math><semantics>𝒞BL<annotation encoding="application / x-tex">\mathcal{C}_{BL}< / annotation>< / semantics>< / com:Math>, of a BL system, which consists of a low SNR region 1004 and a high SNR region 1005, <com:Math><semantics>CTL<annotation encoding="application / x-tex">C_{TL}< / annotation>< / semantics>< / com:Math> in (4) also consists of a low SNR region 1001 and a high SNR region 1003. Unlike BL systems, <com:Math><semantics>𝒞TL<annotation encoding="application / x-tex">\mathcal{C}_{TL}< / annotation>< / semantics>< / com:Math> in (4) also contains a new medium SNR region 1002, when a number, <com:Math><semantics>π<annotation encoding="application / x-tex">\pi< / annotation>< / semantics>< / com:Math>, of the terms, <com:Math><semantics>Λk𝒜K𝔭N0f0<annotation encoding="application / x-tex">\Lambda_k \frac{\mathcal{A}K\mathfrak{p}}{N_0 f_0}< / annotation>< / semantics>< / com:Math>, in (4) are <com:Math><semantics>≪1<annotation encoding="application / x-tex">\ll 1< / annotation>< / semantics>< / com:Math> s.t. <com:Math><semantics>log⁡2(1+𝒜K𝔭N0f0)<annotation encoding="application / x-tex">\log_2 \left(1 + \frac{\mathcal{A}K\mathfrak{p}}{N_0 f_0}\right)< / annotation>< / semantics>< / com:Math> <com:Math><semantics>Λk𝒜K𝔭ℕ0f0≈log⁡2eΛk𝒜K𝔭ℕ0f0<annotation encoding="application / x-tex">\Lambda_k \frac{\mathcal{A}K\mathfrak{p}}{\mathbb{N}_0 f_0} \approx \log_2 e \Lambda_k \frac{\mathcal{A}K\mathfrak{p}}{\mathbb{N}_0 f_0}< / annotation>< / semantics>< / com:Math>. When <com:Math><semantics>𝒞TL<annotation encoding="application / x-tex">\mathcal{C}_{TL}< / annotation>< / semantics>< / com:Math> is in the low SNR region 1001, <com:Math><semantics>𝔫=rank⁡(𝒉Ch)<annotation encoding="application / x-tex">\mathfrak{n} = \operatorname{rank}(\boldsymbol{h}_{Ch})< / annotation>< / semantics>< / com:Math>. When <com:Math><semantics>𝒞TL<annotation encoding="application / x-tex">\mathcal{C}_{TL}< / annotation>< / semantics>< / com:Math> is in the high SNR region 1003, <com:Math><semantics>n=0<annotation encoding="application / x-tex">n = 0< / annotation>< / semantics>< / com:Math>. When <com:Math><semantics>CTL<annotation encoding="application / x-tex">C_{TL}< / annotation>< / semantics>< / com:Math> is in the medium SNR region 1002, <com:Math><semantics>0<n≤1<annotation encoding="application / x-tex">0 < n \le 1< / annotation>< / semantics>< / com:Math> rank <com:Math><semantics>(𝒉Ch)<annotation encoding="application / x-tex">(\boldsymbol{h}_{Ch})< / annotation>< / semantics>< / com:Math>. Given that some of the communication channels in this disclosure are to be constrained by a spectral mask, Theorem I must be modified to include a mask constraint. First, we define the bandwidth (BW) of <com:Math><semantics>xk(t)<annotation encoding="application / x-tex">x_k(t)< / annotation>< / semantics>< / com:Math>, then, introduce the mask constraint. Definition of the BW of <com:Math><semantics>xk(t)<annotation encoding="application / x-tex">x_k(t)< / annotation>< / semantics>< / com:Math> 105, 205, 305: Since <com:Math><semantics>xk(t)<annotation encoding="application / x-tex">x_k(t)< / annotation>< / semantics>< / com:Math> 105, 205, 305 is TL, its PSD, <com:Math><semantics>Sxk(t)(f)<annotation encoding="application / x-tex">S_{x_k(t)}(f)< / annotation>< / semantics>< / com:Math> 802, exists over the entire frequency domain <com:Math><semantics>f∈{−∞,∞}<annotation encoding="application / x-tex">f \in \{-\infty, \infty\}< / annotation>< / semantics>< / com:Math>, allowing for numerous definitions of BW to exist. In this disclosure, we adopt the same definition of BW as the one adopted by the International Telecommunication Union (ITU), which defines transmitter spectrum emissions as falling into three distinct bands: (a) Occupied band 705 emission with a BW <com:Math><semantics>W<annotation encoding="application / x-tex">W< / annotation>< / semantics>< / com:Math>; (b) Out-Of-Band-Emission (OOBE) band 706 with a BW <com:Math><semantics>WOOBE<annotation encoding="application / x-tex">W_{OOBE}< / annotation>< / semantics>< / com:Math>; and (c) Far-Out- Spurious-Emission (FOSE) band 707 with an allowable power level <com:Math><semantics>≤1π<annotation encoding="application / x-tex">\leq \frac{1}{\pi}< / annotation>< / semantics>< / com:Math>. By adopting the same definition for BW as the ITU, we select the BW, <com:Math><semantics>𝒲TL<annotation encoding="application / x-tex">\mathcal{W}_{TL}< / annotation>< / semantics>< / com:Math>, of the TL system to be defined as the BW, <com:Math><semantics>𝒲<annotation encoding="application / x-tex">\mathcal{W}< / annotation>< / semantics>< / com:Math>, of the occupied band 705. Spectral Mask Constraint 701: Some systems considered in this disclosure are constrained by a spectral mask, <com:Math><semantics>ℙMask(f)<annotation encoding="application / x-tex">\mathbb{P}_{Mask}(f)< / annotation>< / semantics>< / com:Math> 701. In this case, <com:Math><semantics>xk(t)<annotation encoding="application / x-tex">x_k(t)< / annotation>< / semantics>< / com:Math> 105, 205, 305 is subject to Constraint 2: Constraint 2: [Image disponible dans le document PDF, Image available in the PDF document] where 7 is a normalization constant, which depends on <com:Math><semantics>𝔭<annotation encoding="application / x-tex">\mathfrak{p}< / annotation>< / semantics>< / com:Math>, <com:Math><semantics>ℙMask(f)<annotation encoding="application / x-tex">\mathbb{P}_{Mask}(f)< / annotation>< / semantics>< / com:Math> 701 and <com:Math><semantics>h<annotation encoding="application / x-tex">h< / annotation>< / semantics>< / com:Math> 1101. According to the ITU, <com:Math><semantics>𝒲<annotation encoding="application / x-tex">\mathcal{W}< / annotation>< / semantics>< / com:Math> must be selected <com:Math><semantics>≤𝒲m<annotation encoding="application / x-tex">\leq \mathcal{W}_m< / annotation>< / semantics>< / com:Math> where <com:Math><semantics>𝒲m<annotation encoding="application / x-tex">\mathcal{W}_m< / annotation>< / semantics>< / com:Math> is the BW of <com:Math><semantics>ℙMask(f)<annotation encoding="application / x-tex">\mathbb{P}_{Mask}(f)< / annotation>< / semantics>< / com:Math> 701. This implies that <com:Math><semantics>𝒲TL<annotation encoding="application / x-tex">\mathcal{W}_{TL}< / annotation>< / semantics>< / com:Math> must be selected <com:Math><semantics>≤𝒲m<annotation encoding="application / x-tex">\leq \mathcal{W}_m< / annotation>< / semantics>< / com:Math> as well. For this reason, we define in this disclosure an overhead factor, <com:Math><semantics>ψ≜𝒲TLNTs≥1<annotation encoding="application / x-tex">\psi \triangleq \mathcal{W}_{TL}NT_s \geq 1< / annotation>< / semantics>< / com:Math>, as the overhead, both in time and in frequency, which is required for <com:Math><semantics>xk(t)<annotation encoding="application / x-tex">x_k(t)< / annotation>< / semantics>< / com:Math> 105, 205, 305 to comply with Constraint 2. It is selected such that <com:Math><semantics>𝒲TL≤𝒲m<annotation encoding="application / x-tex">\mathcal{W}_{TL} \leq \mathcal{W}_m< / annotation>< / semantics>< / com:Math> or equivalently, N is selected s.t. <com:Math><semantics>N≥Nmin≜rWmTs<annotation encoding="application / x-tex">N \ge N_{min} \triangleq \frac{r}{W_m T_s}< / annotation>< / semantics>< / com:Math>. Under Constraints 1-2, <com:Math><semantics>CTL<annotation encoding="application / x-tex">C_{TL}< / annotation>< / semantics>< / com:Math> in (4) can be expressed as [Image disponible dans le document PDF, Image available in the PDF document] bps (5) Similarly, under Constraints 1-2, a BL system, of fixed BW, <com:Math><semantics>WBL<annotation encoding="application / x-tex">W_{BL}< / annotation>< / semantics>< / com:Math>, selected as <com:Math><semantics>WBL=Wm<annotation encoding="application / x-tex">W_{BL} = W_m< / annotation>< / semantics>< / com:Math>, has a BL capacity, <com:Math><semantics>CBL<annotation encoding="application / x-tex">C_{BL}< / annotation>< / semantics>< / com:Math>, given as [Image disponible dans le document PDF, Image available in the PDF document] (6) where <com:Math><semantics>𝒱BL<annotation encoding="application / x-tex">\mathcal{V}_{BL}< / annotation>< / semantics>< / com:Math> is defined as the overhead factor, both in time and in frequency, which is required for the BL system to comply with Constraint 2. When <com:Math><semantics>𝒜‾K𝔭Nowm>1<annotation encoding="application / x-tex">\frac{\bar{\mathcal{A}}K\mathfrak{p}}{N_o w_m} > 1< / annotation>< / semantics>< / com:Math>, (6) implies that doubling <com:Math><semantics>𝒞BL<annotation encoding="application / x-tex">\mathcal{C}_{BL}< / annotation>< / semantics>< / com:Math>, with a fixed BW, requires a geometric multiple increase in <com:Math><semantics>𝒜‾K𝔭ℕo𝒲m<annotation encoding="application / x-tex">\frac{\bar{\mathcal{A}}K\mathfrak{p}}{\mathbb{N}_o \mathcal{W}_m}< / annotation>< / semantics>< / com:Math> since its contribution towards <com:Math><semantics>𝒞BL<annotation encoding="application / x-tex">\mathcal{C}_{BL}< / annotation>< / semantics>< / com:Math> is logarithmic. FIG. 10a compares <com:Math><semantics>𝒞TL / 𝒲m<annotation encoding="application / x-tex">\mathcal{C}_{TL} / \mathcal{W}_m< / annotation>< / semantics>< / com:Math> in (5) (shown with "." markers) with <com:Math><semantics>𝒞BL / 𝒲m<annotation encoding="application / x-tex">\mathcal{C}_{BL} / \mathcal{W}_m< / annotation>< / semantics>< / com:Math> in (6) (shown with "*" markers) versus the average received SNR that is normalized w.r.t. <com:Math><semantics>𝒲m<annotation encoding="application / x-tex">\mathcal{W}_m< / annotation>< / semantics>< / com:Math>, when <com:Math><semantics>r=rBL=<annotation encoding="application / x-tex">r = r_{BL} =< / annotation>< / semantics>< / com:Math> 1, <com:Math><semantics>d=K=1<annotation encoding="application / x-tex">d = K = 1< / annotation>< / semantics>< / com:Math> and the <com:Math><semantics>kth<annotation encoding="application / x-tex">k^{th}< / annotation>< / semantics>< / com:Math> column, <com:Math><semantics>{hCh}k<annotation encoding="application / x-tex">\{h_{Ch}\}_k< / annotation>< / semantics>< / com:Math> of <com:Math><semantics>hCh<annotation encoding="application / x-tex">h_{Ch}< / annotation>< / semantics>< / com:Math> corresponds to a rectangular pulse <com:Math><semantics>∀k<annotation encoding="application / x-tex">\forall k< / annotation>< / semantics>< / com:Math>. In FIG. 10a, <com:Math><semantics>CTL / Wm<annotation encoding="application / x-tex">C_{TL} / W_m< / annotation>< / semantics>< / com:Math> in (5) is illustrated as a number of curves, each curve corresponding to a value of <com:Math><semantics>Nmin<annotation encoding="application / x-tex">N_{min}< / annotation>< / semantics>< / com:Math>. The selected values are <com:Math><semantics>Nmin=1,2,4,...,1024<annotation encoding="application / x-tex">N_{min}=1,2,4,...,1024< / annotation>< / semantics>< / com:Math>, with <com:Math><semantics>Nmin=1<annotation encoding="application / x-tex">N_{min}=1< / annotation>< / semantics>< / com:Math> coinciding with <com:Math><semantics>CBL / Wm<annotation encoding="application / x-tex">C_{BL} / W_m< / annotation>< / semantics>< / com:Math> in (6). Similar to <com:Math><semantics>CBL / Wm<annotation encoding="application / x-tex">C_{BL} / W_m< / annotation>< / semantics>< / com:Math> in (6), <com:Math><semantics>CTL / Wm<annotation encoding="application / x-tex">C_{TL} / W_m< / annotation>< / semantics>< / com:Math> in (5) contains a "Low" SNR region 1001 and a "High" SNR region 1003. Unlike <com:Math><semantics>𝒞BL / 𝒲m<annotation encoding="application / x-tex">\mathcal{C}_{BL} / \mathcal{W}_m< / annotation>< / semantics>< / com:Math> in (6), which contains a low SNR region 1004, and a high SNR region 1005, <com:Math><semantics>CTL / Wm<annotation encoding="application / x-tex">C_{TL} / W_m< / annotation>< / semantics>< / com:Math> in (5) also contains a medium SNR region 1002, denoted as "Med. SNR" in FIG. 10a where doubling <com:Math><semantics>𝒞TL / 𝒲m<annotation encoding="application / x-tex">\mathcal{C}_{TL} / \mathcal{W}_m< / annotation>< / semantics>< / com:Math> in (5) requires a fixed multiple increase in <com:Math><semantics>ΛkrNmin𝒜K𝔭ℕo𝒲m<annotation encoding="application / x-tex">\frac{\Lambda_k r}{N_{min}} \frac{\mathcal{A}K\mathfrak{p}}{\mathbb{N}_o \mathcal{W}_m}< / annotation>< / semantics>< / com:Math> since its contribution towards <com:Math><semantics>𝒞TL<annotation encoding="application / x-tex">\mathcal{C}_{TL}< / annotation>< / semantics>< / com:Math> is mostly linear. Interpretation of FIG. 10a: The medium SNR region 1002 is created in <com:Math><semantics>𝒞TL<annotation encoding="application / x-tex">\mathcal{C}_{TL}< / annotation>< / semantics>< / com:Math> in (5) when the average received BL SNR, <com:Math><semantics>𝒜‾K𝔭Nowm<annotation encoding="application / x-tex">\frac{\bar{\mathcal{A}}K\mathfrak{p}}{N_o w_m}< / annotation>< / semantics>< / com:Math>, is <com:Math><semantics>>1<annotation encoding="application / x-tex">> 1< / annotation>< / semantics>< / com:Math> while <com:Math><semantics>𝒜‾K𝔭NowmΛkrNmin<annotation encoding="application / x-tex">\frac{\bar{\mathcal{A}}K\mathfrak{p}}{N_o w_m} \frac{\Lambda_k r}{N_{min}}< / annotation>< / semantics>< / com:Math> is <com:Math><semantics>≪1<annotation encoding="application / x-tex">\ll 1< / annotation>< / semantics>< / com:Math>. In other words, <com:Math><semantics>ΛkrNmin<annotation encoding="application / x-tex">\frac{\Lambda_k r}{N_{min}}< / annotation>< / semantics>< / com:Math> must be much smaller than <com:Math><semantics>𝒜‾K𝔭ℕo𝒲m<annotation encoding="application / x-tex">\frac{\bar{\mathcal{A}}K\mathfrak{p}}{\mathbb{N}_o \mathcal{W}_m}< / annotation>< / semantics>< / com:Math> in order to create the medium SNR region 1002. The source for having <com:Math><semantics>Λk<annotation encoding="application / x-tex">\Lambda_k< / annotation>< / semantics>< / com:Math> small and <com:Math><semantics>Nmin<annotation encoding="application / x-tex">N_{min}< / annotation>< / semantics>< / com:Math> large while keeping <com:Math><semantics>ψ<∞<annotation encoding="application / x-tex">\psi < \infty< / annotation>< / semantics>< / com:Math> is having an arbitrarily large number of DOF, while complying with Constraint 2. In a practical design requiring finite latency, all DOF must have a Finite Access Time (FAT), or equivalently, the time it takes to access any such DOF is finite. We refer to such DOF as FAT, and observe that only TL systems have an arbitrarily large number of FAT DOF in their high frequency components, while BL systems have only a finite number of FAT DOF since they are not allowed to contain high frequency components. Attribute of h for <com:Math><semantics>∼<annotation encoding="application / x-tex">\sim< / annotation>< / semantics>< / com:Math> to be <com:Math><semantics><∞<annotation encoding="application / x-tex">< \infty< / annotation>< / semantics>< / com:Math>: Given that some of the communication channels 106, 306, 406 in this disclosure are to be constrained by a spectral mask 701 with <com:Math><semantics>≪∞<annotation encoding="application / x-tex">\ll \infty< / annotation>< / semantics>< / com:Math>, it is imperative to analyze <com:Math><semantics>Sxk(t)(f)<annotation encoding="application / x-tex">S_{x_k(t)}(f)< / annotation>< / semantics>< / com:Math> 801. An important attribute of <com:Math><semantics>h<annotation encoding="application / x-tex">h< / annotation>< / semantics>< / com:Math> 1101, which affects the spectral decay of <com:Math><semantics>Sxk(t)(f)<annotation encoding="application / x-tex">S_{x_k(t)}(f)< / annotation>< / semantics>< / com:Math> 801, is the Degree of Differenciability (DOD), <com:Math><semantics>ℵk<annotation encoding="application / x-tex">\aleph_k< / annotation>< / semantics>< / com:Math>, of the <com:Math><semantics>kth<annotation encoding="application / x-tex">k^{th}< / annotation>< / semantics>< / com:Math> column, <com:Math><semantics>{h}k<annotation encoding="application / x-tex">\{h\}_k< / annotation>< / semantics>< / com:Math> of <com:Math><semantics>h<annotation encoding="application / x-tex">h< / annotation>< / semantics>< / com:Math> 1101, defined as the number of times <com:Math><semantics>{h}k<annotation encoding="application / x-tex">\{h\}_k< / annotation>< / semantics>< / com:Math> can be differenced in time until a Dirac delta impulse, <com:Math><semantics>δ<annotation encoding="application / x-tex">\delta< / annotation>< / semantics>< / com:Math>, appears. Mathematically, this implies that <com:Math><semantics>ℵ≜min⁡{Δ→kn}i=δ{n}<annotation encoding="application / x-tex">\aleph \triangleq \min_{\{\vec{\Delta}_k^n\}_i = \delta} \{n\}< / annotation>< / semantics>< / com:Math> where <com:Math><semantics>{Δ→kn}l<annotation encoding="application / x-tex">\{\vec{\Delta}_k^n\}_l< / annotation>< / semantics>< / com:Math> is the <com:Math><semantics>lth<annotation encoding="application / x-tex">l^{th}< / annotation>< / semantics>< / com:Math> element in the differencing vector, <com:Math><semantics>Δ→kn<annotation encoding="application / x-tex">\vec{\Delta}_k^n< / annotation>< / semantics>< / com:Math>, of order n, corresponding to <com:Math><semantics>{h}k<annotation encoding="application / x-tex">\{h\}_k< / annotation>< / semantics>< / com:Math>, and defined as [Image disponible dans le document PDF, Image available in the PDF document] with initial condition: <com:Math><semantics>{Δ¯k0}l≜{h}k,l,l∈{1,...,N−1}.<annotation encoding="application / x-tex">\{\overline{\Delta}_k^0\}_l \triangleq \{h\}_{k,l}, l \in \{1, ..., N-1\}.< / annotation>< / semantics>< / com:Math> Examples of X: • When <com:Math><semantics>{h}k<annotation encoding="application / x-tex">\{h\}_k< / annotation>< / semantics>< / com:Math> is a TL rectangular pulse, <com:Math><semantics>ℵk=1<annotation encoding="application / x-tex">\aleph_k = 1< / annotation>< / semantics>< / com:Math>. • When <com:Math><semantics>{h}k<annotation encoding="application / x-tex">\{h\}_k< / annotation>< / semantics>< / com:Math> is one lobe of a sine wave, <com:Math><semantics>ℵk=2<annotation encoding="application / x-tex">\aleph_k = 2< / annotation>< / semantics>< / com:Math>. • When <com:Math><semantics>{h}k<annotation encoding="application / x-tex">\{h\}_k< / annotation>< / semantics>< / com:Math> is a pseudo-noise (PN) sequence, <com:Math><semantics>ℵk=0<annotation encoding="application / x-tex">\aleph_k = 0< / annotation>< / semantics>< / com:Math>. The following 2 DOD properties are used below: • DOD Property I: When <com:Math><semantics>{h}k<annotation encoding="application / x-tex">\{h\}_k< / annotation>< / semantics>< / com:Math> is the sum of two TL vectors, <com:Math><semantics>{h1}k<annotation encoding="application / x-tex">\{h_1\}_k< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>{h2}k<annotation encoding="application / x-tex">\{h_2\}_k< / annotation>< / semantics>< / com:Math>, i.e. <com:Math><semantics>{h}k=<annotation encoding="application / x-tex">\{h\}_k =< / annotation>< / semantics>< / com:Math> <com:Math><semantics>{h1}k+{h2}k<annotation encoding="application / x-tex">\{h_1\}_k + \{h_2\}_k< / annotation>< / semantics>< / com:Math>, with respective DOD, <com:Math><semantics>ℵk,1<annotation encoding="application / x-tex">\aleph_{k,1}< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>ℵk,2<annotation encoding="application / x-tex">\aleph_{k,2}< / annotation>< / semantics>< / com:Math>, its resulting DOD, <com:Math><semantics>ℵk<annotation encoding="application / x-tex">\aleph_k< / annotation>< / semantics>< / com:Math>, is asymptotically equal to <com:Math><semantics>lim⁡N→∞ℵk=min⁡(ℵk,1,ℵk,2)<annotation encoding="application / x-tex">\lim_{N\to\infty} \aleph_k = \min(\aleph_{k,1}, \aleph_{k,2})< / annotation>< / semantics>< / com:Math>. • DOD Property II: When <com:Math><semantics>{h}k<annotation encoding="application / x-tex">\{h\}_k< / annotation>< / semantics>< / com:Math> is the linear or circular convolution between two TL vectors, <com:Math><semantics>{h1}k<annotation encoding="application / x-tex">\{h_1\}_k< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>{h2}k<annotation encoding="application / x-tex">\{h_2\}_k< / annotation>< / semantics>< / com:Math> with respective DOD, <com:Math><semantics>ℵk,1<annotation encoding="application / x-tex">\aleph_{k,1}< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>ℵk,2<annotation encoding="application / x-tex">\aleph_{k,2}< / annotation>< / semantics>< / com:Math>, its resulting DOD, <com:Math><semantics>ℵk<annotation encoding="application / x-tex">\aleph_k< / annotation>< / semantics>< / com:Math>, is: <com:Math><semantics>ℵk=ℵk,1+ℝ<annotation encoding="application / x-tex">\aleph_k = \aleph_{k,1} + \mathbb{R}< / annotation>< / semantics>< / com:Math> <com:Math><semantics>ℵk,2<annotation encoding="application / x-tex">\aleph_{k,2}< / annotation>< / semantics>< / com:Math> Theorem II derives the slope of the medium SNR region 1002 as a function of the DOD of <com:Math><semantics>hCh<annotation encoding="application / x-tex">h_{Ch}< / annotation>< / semantics>< / com:Math>. Theorem II: Doubling <com:Math><semantics>CTL<annotation encoding="application / x-tex">C_{TL}< / annotation>< / semantics>< / com:Math> in (4) across its medium SNR region 1002 requires increasing <com:Math><semantics>𝒜‾K𝔭ℕofs<annotation encoding="application / x-tex">\frac{\bar{\mathcal{A}}K\mathfrak{p}}{\mathbb{N}_o f_s}< / annotation>< / semantics>< / com:Math> by a fixed multiple of <com:Math><semantics>22ℵk<annotation encoding="application / x-tex">2^{2\aleph_k}< / annotation>< / semantics>< / com:Math> where <com:Math><semantics>ℵk<annotation encoding="application / x-tex">\aleph_k< / annotation>< / semantics>< / com:Math> is the DOD of <com:Math><semantics>hCh<annotation encoding="application / x-tex">h_{Ch}< / annotation>< / semantics>< / com:Math>. The following constraint derives the modulation, which maximizes n, when using a Minimum Mean Square Error with Successive Interference Cancellation (MMSE-SIC) detector 110, 210, 310 at Rx 108, 208, 308, selected for its low complexity and its asymptotic optimality under certain conditions. This constraint maximizes <com:Math><semantics>CTL<annotation encoding="application / x-tex">C_{TL}< / annotation>< / semantics>< / com:Math> in the medium SNR region 1002. Modulation Constraint: It is possible to show that minimizing the arithmetic mean of the MMSE at Rx 108, 208, 308 is equivalent to maximizing <com:Math><semantics>ddSNRνlog⁡2(1+SNRkvν)∀k<annotation encoding="application / x-tex">\frac{d}{d SNR_{\nu}} \log_2 \left(1 + \frac{SNR_k}{v_{\nu}}\right) \forall k< / annotation>< / semantics>< / com:Math> where <com:Math><semantics>SNRk<annotation encoding="application / x-tex">SNR_k< / annotation>< / semantics>< / com:Math> is the received normalized SNR corresponding to <com:Math><semantics>{hCh}k<annotation encoding="application / x-tex">\{h_{Ch}\}_k< / annotation>< / semantics>< / com:Math> while <com:Math><semantics>1pk<annotation encoding="application / x-tex">\frac{1}{p_k}< / annotation>< / semantics>< / com:Math> is its multiuser efficiency. Unlike water-filling, which deals with parallel channels, the solution for such optimization is <com:Math><semantics>SNRkdt=1∀k<annotation encoding="application / x-tex">\frac{SNR_k}{dt} = 1 \ \forall k< / annotation>< / semantics>< / com:Math>. This implies that the modulation of choice for the elements of <com:Math><semantics>α→<annotation encoding="application / x-tex">\vec{\alpha}< / annotation>< / semantics>< / com:Math> corresponds to <com:Math><semantics>pk<annotation encoding="application / x-tex">p_k< / annotation>< / semantics>< / com:Math> loading each DOF with about 1 bit of information. In comparison, the low SNR regions 1001, 1004 correspond to loading < 1 bit / DOF, while the high SNR regions 1003, 1005 typically correspond to loading > 1 bit / DOF. When <com:Math><semantics>Ak≠A<annotation encoding="application / x-tex">A_k \neq A< / annotation>< / semantics>< / com:Math> for some k, we use instead: Constraint 3: [Image disponible dans le document PDF, Image available in the PDF document] Theorem III modifies Theorem I to include a mask constraint 701 and a modulation constraint. Theorem III: The capacity, <com:Math><semantics>CTL<annotation encoding="application / x-tex">C_{TL}< / annotation>< / semantics>< / com:Math>, of the TL channel corresponding to <com:Math><semantics>hCh<annotation encoding="application / x-tex">h_{Ch}< / annotation>< / semantics>< / com:Math> in (3), subject to Constraints 1-3, with <com:Math><semantics>K≤d<annotation encoding="application / x-tex">K \le d< / annotation>< / semantics>< / com:Math> and with <com:Math><semantics>ℵk>0<annotation encoding="application / x-tex">\aleph_k > 0< / annotation>< / semantics>< / com:Math> is [Image disponible dans le document PDF, Image available in the PDF document] (7) where <com:Math><semantics>pk<annotation encoding="application / x-tex">p_k< / annotation>< / semantics>< / com:Math> is asymptotically <com:Math><semantics>∝1rK2ℵk−1Nd2ℵk<annotation encoding="application / x-tex">\propto \frac{1}{r} K^{2\aleph_k - 1} N_d^{2\aleph_k}< / annotation>< / semantics>< / com:Math> as <com:Math><semantics>N≫1<annotation encoding="application / x-tex">N \gg 1< / annotation>< / semantics>< / com:Math> with <com:Math><semantics>Nd≜⌈N / d⌉<annotation encoding="application / x-tex">N_d \triangleq \lceil N / d \rceil< / annotation>< / semantics>< / com:Math>, the ceiling of <com:Math><semantics>N / d<annotation encoding="application / x-tex">N / d< / annotation>< / semantics>< / com:Math>, and <com:Math><semantics>N≥Nmin<annotation encoding="application / x-tex">N \ge N_{min}< / annotation>< / semantics>< / com:Math>, using an MMSE-SIC detector 110, 210, 310 at Rx 108, 208, 308. Importance of Theorem III: In FIG. 10a <com:Math><semantics>CTL / Wm<annotation encoding="application / x-tex">C_{TL} / W_m< / annotation>< / semantics>< / com:Math> in (7) is illustrated as a number of points (shown with "square" markers), each corresponding to a value of <com:Math><semantics>Nmin<annotation encoding="application / x-tex">N_{min}< / annotation>< / semantics>< / com:Math>. FIG. 10a confirms that Constraint 3 maximizes <com:Math><semantics>𝒞TL<annotation encoding="application / x-tex">\mathcal{C}_{TL}< / annotation>< / semantics>< / com:Math> in the medium SNR region. Based on (7) and on Constraint 3, <com:Math><semantics>lim⁡L→∞𝒞TL≥wmrNKd<annotation encoding="application / x-tex">\lim_{L\to\infty} \mathcal{C}_{TL} \geq \frac{w_m}{r} \frac{NK}{d}< / annotation>< / semantics>< / com:Math>. Therefore, doubling <com:Math><semantics>lim⁡L→∞𝒞TL<annotation encoding="application / x-tex">\lim_{L\to\infty} \mathcal{C}_{TL}< / annotation>< / semantics>< / com:Math> by doubling <com:Math><semantics>Nd<annotation encoding="application / x-tex">\frac{N}{d}< / annotation>< / semantics>< / com:Math>, for a fixed <com:Math><semantics>wmr<annotation encoding="application / x-tex">\frac{w_m}{r}< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>K<annotation encoding="application / x-tex">K< / annotation>< / semantics>< / com:Math>, requires increasing <com:Math><semantics>SNRk<annotation encoding="application / x-tex">SNR_k< / annotation>< / semantics>< / com:Math> by a fixed multiple of <com:Math><semantics>22ℵk<annotation encoding="application / x-tex">2^{2\aleph_k}< / annotation>< / semantics>< / com:Math>, while doubling <com:Math><semantics>lim⁡L→∞𝒞TL<annotation encoding="application / x-tex">\lim_{L\to\infty} \mathcal{C}_{TL}< / annotation>< / semantics>< / com:Math> by doubling <com:Math><semantics>K<annotation encoding="application / x-tex">K< / annotation>< / semantics>< / com:Math>, for a fixed <com:Math><semantics>wmr<annotation encoding="application / x-tex">\frac{w_m}{r}< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>Nd<annotation encoding="application / x-tex">\frac{N}{d}< / annotation>< / semantics>< / com:Math>, requires increasing <com:Math><semantics>SNRk<annotation encoding="application / x-tex">SNR_k< / annotation>< / semantics>< / com:Math> by a fixed multiple of <com:Math><semantics>22ℵk−1<annotation encoding="application / x-tex">2^{2\aleph_k-1}< / annotation>< / semantics>< / com:Math> when <com:Math><semantics>K≤d<annotation encoding="application / x-tex">K \leq d< / annotation>< / semantics>< / com:Math>. The next section introduces a novel TL system with FAT DOF, referred to as an MTF system. 5.2 MTF DESIGN Design Problem: h 1101 in (1) is to be designed based on Theorems I and III with the goal of achieving a desired channel capacity, <com:Math><semantics>𝒞b<annotation encoding="application / x-tex">\mathcal{C}_{b}< / annotation>< / semantics>< / com:Math> 1301, for a given channel 106, 306, 406 of BW <com:Math><semantics>𝒲m<annotation encoding="application / x-tex">\mathcal{W}_{m}< / annotation>< / semantics>< / com:Math>. Three design steps, MTF Design Steps I-III 1302, 1304, 1306, are shown below, followed by a proposed MTF design implementation. All 3 steps attempt to design h 1101 such that the minimum required average received SNR is minimized for a given desired capacity, <com:Math><semantics>Ch<annotation encoding="application / x-tex">C_h< / annotation>< / semantics>< / com:Math> 1301, and for a given BW <com:Math><semantics>𝒲m<annotation encoding="application / x-tex">\mathcal{W}_m< / annotation>< / semantics>< / com:Math>. This requires designing <com:Math><semantics>h<annotation encoding="application / x-tex">h< / annotation>< / semantics>< / com:Math> 1101 s.t. the set, <com:Math><semantics>{Λk}k=1rank⁡(hCh)<annotation encoding="application / x-tex">\{\Lambda_k\}_{k=1}^{\operatorname{rank}(h_{Ch})}< / annotation>< / semantics>< / com:Math>, of squared singular values of <com:Math><semantics>hCh<annotation encoding="application / x-tex">h_{Ch}< / annotation>< / semantics>< / com:Math> in (3) has a variance that is minimized while complying with Constraint 2. MTF Design Solution: First, we define <com:Math><semantics>Ts,max≜KdCb≥1 / fs<annotation encoding="application / x-tex">T_{s,max} \triangleq \frac{K}{dC_b} \geq 1 / f_s< / annotation>< / semantics>< / com:Math> 1301 as a function of K and <com:Math><semantics>d<annotation encoding="application / x-tex">d< / annotation>< / semantics>< / com:Math>, which depend on the selected TL channel. For example, when the TL channel has relatively low interference, such as with <com:Math><semantics>K=1<annotation encoding="application / x-tex">K = 1< / annotation>< / semantics>< / com:Math>, one can select the TL system to be with memory, i.e. with <com:Math><semantics>d=1<N<annotation encoding="application / x-tex">d = 1 < N< / annotation>< / semantics>< / com:Math>, implying that <com:Math><semantics>Ts,max=1cs≥1 / fs<annotation encoding="application / x-tex">T_{s,max} = \frac{1}{c_s} \ge 1 / f_s< / annotation>< / semantics>< / com:Math> or <com:Math><semantics>fs≥cb<annotation encoding="application / x-tex">f_s \ge c_b< / annotation>< / semantics>< / com:Math> 1301. On the other hand, when the TL channel has relatively high interference, i.e. with <com:Math><semantics>K≫1<annotation encoding="application / x-tex">K \gg 1< / annotation>< / semantics>< / com:Math>, one can select the TL system to be memoryless, i.e. with <com:Math><semantics>d=N<annotation encoding="application / x-tex">d = N< / annotation>< / semantics>< / com:Math>, implying that <com:Math><semantics>Ts,max=KNCb≥1 / fs<annotation encoding="application / x-tex">T_{s,max} = \frac{K}{N C_b} \ge 1 / f_s< / annotation>< / semantics>< / com:Math> or <com:Math><semantics>fs≥NKCb<annotation encoding="application / x-tex">f_s \ge \frac{N}{K} C_b< / annotation>< / semantics>< / com:Math> 1301. MTF Design Step I 1302: For a fixed <com:Math><semantics>Ts≤Ts,max<annotation encoding="application / x-tex">T_s \leq T_{s,max}< / annotation>< / semantics>< / com:Math> 1301, select the number, N 1303, of FAT DOF as <com:Math><semantics>N≜NS+NJ<annotation encoding="application / x-tex">N \triangleq N_S + N_J< / annotation>< / semantics>< / com:Math>, where: a) <com:Math><semantics>NS<annotation encoding="application / x-tex">N_S< / annotation>< / semantics>< / com:Math> is defined as the number of Shaping FAT (S-FAT) DOF 1303 selected such that <com:Math><semantics>NS≥<annotation encoding="application / x-tex">N_S \ge< / annotation>< / semantics>< / com:Math> <com:Math><semantics>Nmin<annotation encoding="application / x-tex">N_{min}< / annotation>< / semantics>< / com:Math> in order to comply with the BW constraint, <com:Math><semantics>𝒲TL≤𝒲m<annotation encoding="application / x-tex">\mathcal{W}_{TL} \leq \mathcal{W}_m< / annotation>< / semantics>< / com:Math>, of <com:Math><semantics>ℙMask(f)<annotation encoding="application / x-tex">\mathbb{P}_{Mask}(f)< / annotation>< / semantics>< / com:Math> 701; and b) <com:Math><semantics>NJ<annotation encoding="application / x-tex">N_J< / annotation>< / semantics>< / com:Math> is defined as the number of Interpolating FAT (I-FAT) DOF 1303 obtained through the creation of interpolated sampled frequencies 1303 inside the existing occupied band 705. Power is taken from existing frequencies and allocated to the newly formed frequencies 1303 s.t. Constraint 1 is preserved. Commonly, <com:Math><semantics>Nℐ=0<annotation encoding="application / x-tex">N_{\mathcal{I}} = 0< / annotation>< / semantics>< / com:Math> since no constraints depend on <com:Math><semantics>Nℐ<annotation encoding="application / x-tex">N_{\mathcal{I}}< / annotation>< / semantics>< / com:Math>. MTF Design Step II 1304: Once N 1303 is selected and the newly sampled frequencies are created 1303, <com:Math><semantics>pk<annotation encoding="application / x-tex">p_k< / annotation>< / semantics>< / com:Math> can be reduced by equalizing the power, <com:Math><semantics>E{|Hk(Ω)|2}<annotation encoding="application / x-tex">E\{|H_k(\Omega)|^2\}< / annotation>< / semantics>< / com:Math>, across <com:Math><semantics>Ω∈[−π,π]<annotation encoding="application / x-tex">\Omega \in [-\pi, \pi]< / annotation>< / semantics>< / com:Math> as much as possible, while preserving Constraint 2, where <com:Math><semantics>Hk(Ω)<annotation encoding="application / x-tex">H_k(\Omega)< / annotation>< / semantics>< / com:Math> is the Discrete-Time Fourier Transform (DTFT) of the <com:Math><semantics>kth<annotation encoding="application / x-tex">k^{th}< / annotation>< / semantics>< / com:Math> column, <com:Math><semantics>{hBasic}k<annotation encoding="application / x-tex">\{h_{Basic}\}_k< / annotation>< / semantics>< / com:Math>, of <com:Math><semantics>hBasic<annotation encoding="application / x-tex">h_{Basic}< / annotation>< / semantics>< / com:Math> 1102, 1103 and <com:Math><semantics>Ω<annotation encoding="application / x-tex">\Omega< / annotation>< / semantics>< / com:Math> is the normalized frequency. This equalization is defined as taking power from frequency samples 1303 with above average power and allocating it to frequencies 1305 with below average power, thereby preserving Constraint 1. It can be shown using Karamata's inequality that such power allocation reduces the variance of <com:Math><semantics>{Λk}k=1rank⁡(hCh)<annotation encoding="application / x-tex">\{\Lambda_k\}_{k=1}^{\operatorname{rank}(h_{Ch})}< / annotation>< / semantics>< / com:Math> hence increasing <com:Math><semantics>𝒞TL<annotation encoding="application / x-tex">\mathcal{C}_{TL}< / annotation>< / semantics>< / com:Math>. MTF Design Step III 1306: Once N 1303 is selected, the newly sampled frequencies 1303 are created, and the power, <com:Math><semantics>E{|Hk(Ω)|2}<annotation encoding="application / x-tex">E\{|H_k(\Omega)|^2\}< / annotation>< / semantics>< / com:Math>, across <com:Math><semantics>Ω∈[−π,π]<annotation encoding="application / x-tex">\Omega \in [-\pi, \pi]< / annotation>< / semantics>< / com:Math> is equalized 1305 as much as possible, <com:Math><semantics>pk<annotation encoding="application / x-tex">p_k< / annotation>< / semantics>< / com:Math> can be reduced by selecting the phases of the samples of <com:Math><semantics>Hk(Ω)<annotation encoding="application / x-tex">H_k(\Omega)< / annotation>< / semantics>< / com:Math>, s.t. the entries 1307 of <com:Math><semantics>hBasic<annotation encoding="application / x-tex">h_{Basic}< / annotation>< / semantics>< / com:Math> 1102, 1103 are zero mean RVs, (ideally) Gaussian. This assignment of the phases in <com:Math><semantics>Hk(Ω)<annotation encoding="application / x-tex">H_k(\Omega)< / annotation>< / semantics>< / com:Math> does not affect the power, <com:Math><semantics>E{|Hk(Ω)|2}<annotation encoding="application / x-tex">E\{|H_k(\Omega)|^2\}< / annotation>< / semantics>< / com:Math>, across <com:Math><semantics>Ω∈[−π,π]<annotation encoding="application / x-tex">\Omega \in [-\pi, \pi]< / annotation>< / semantics>< / com:Math>, and thus, preserves Constraints 1-2. Nomenclature: We refer to h 1101 designed based on MTF Design Steps I-III 1302, 1304, 1306, and subject to Constraints 1-3, as an MTF matrix. In this case, we denote <com:Math><semantics>h<annotation encoding="application / x-tex">h< / annotation>< / semantics>< / com:Math> 1101 as <com:Math><semantics>hMTF<annotation encoding="application / x-tex">h_{MTF}< / annotation>< / semantics>< / com:Math> 1104, <com:Math><semantics>hCh<annotation encoding="application / x-tex">h_{Ch}< / annotation>< / semantics>< / com:Math> as <com:Math><semantics>hMTF,Ch<annotation encoding="application / x-tex">h_{MTF,Ch}< / annotation>< / semantics>< / com:Math>, <com:Math><semantics>hBasic<annotation encoding="application / x-tex">h_{Basic}< / annotation>< / semantics>< / com:Math> 1102, 1103 as <com:Math><semantics>hB_MTF<annotation encoding="application / x-tex">h_{B\_MTF}< / annotation>< / semantics>< / com:Math>, <com:Math><semantics>CTL<annotation encoding="application / x-tex">C_{TL}< / annotation>< / semantics>< / com:Math> in (7) as <com:Math><semantics>CMTF<annotation encoding="application / x-tex">C_{MTF}< / annotation>< / semantics>< / com:Math> and refer to the combination of the MTF system and of the channel as the MTF channel. <com:Math><semantics>hBMTF<annotation encoding="application / x-tex">h_{B\ MTF}< / annotation>< / semantics>< / com:Math> is defined by the building blocks <com:Math><semantics>hB_MTF,1<annotation encoding="application / x-tex">h_{B\_MTF,1}< / annotation>< / semantics>< / com:Math> 1105,..., <com:Math><semantics>hBMTF,Nd<annotation encoding="application / x-tex">h_{B_{MTF},N_d}< / annotation>< / semantics>< / com:Math> 1106. The <com:Math><semantics>kth<annotation encoding="application / x-tex">k^{th}< / annotation>< / semantics>< / com:Math> column <com:Math><semantics>{hB_MTF}ν<annotation encoding="application / x-tex">\{h_{B\_MTF}\}_{\nu}< / annotation>< / semantics>< / com:Math> 1308 of <com:Math><semantics>hBMTF<annotation encoding="application / x-tex">h_{B\ MTF}< / annotation>< / semantics>< / com:Math> is obtained from <com:Math><semantics>Hk(Ω)<annotation encoding="application / x-tex">H_k(\Omega)< / annotation>< / semantics>< / com:Math> using an inverse DTFT 1310. MTF Design Implementation: An implementation of MTF Design Steps I-III 1302, 1304, 1306 is proposed here where the <com:Math><semantics>kth<annotation encoding="application / x-tex">k^{th}< / annotation>< / semantics>< / com:Math> column, <com:Math><semantics>{hB_MTF}k<annotation encoding="application / x-tex">\{h_{B\_MTF}\}_k< / annotation>< / semantics>< / com:Math> 1308, of <com:Math><semantics>hB_MTF<annotation encoding="application / x-tex">h_{B\_MTF}< / annotation>< / semantics>< / com:Math>, is expressed as a sum: [Image disponible dans le document PDF, Image available in the PDF document] (8) of 2 vectors, <com:Math><semantics>℘→k,ℵk<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,\aleph_k}< / annotation>< / semantics>< / com:Math> 1208 and <com:Math><semantics>℘→k,0<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,0}< / annotation>< / semantics>< / com:Math> 1209, defined as follows: Vector I: <com:Math><semantics>℘→k,ℵk∈ℂN×1<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,\aleph_k} \in \mathbb{C}^{N\times 1}< / annotation>< / semantics>< / com:Math> 1208 is a pulse vector with a DOD, <com:Math><semantics>ℵk>0<annotation encoding="application / x-tex">\aleph_k > 0< / annotation>< / semantics>< / com:Math>, selected in order for <com:Math><semantics>{hMTF}k<annotation encoding="application / x-tex">\{h_{MTF}\}_k< / annotation>< / semantics>< / com:Math> to comply with the BW constraint, i.e. <com:Math><semantics>𝒲TL≤𝒲m<annotation encoding="application / x-tex">\mathcal{W}_{TL} \leq \mathcal{W}_m< / annotation>< / semantics>< / com:Math> 705. It is formed using <com:Math><semantics>ℵk<annotation encoding="application / x-tex">\aleph_k< / annotation>< / semantics>< / com:Math> linear convolutions (each denoted by '*' 1207): [Image disponible dans le document PDF, Image available in the PDF document] (9) between <com:Math><semantics>ℵk+1>0<annotation encoding="application / x-tex">\aleph_k + 1 > 0< / annotation>< / semantics>< / com:Math> vectors, with the <com:Math><semantics>lth<annotation encoding="application / x-tex">l^{th}< / annotation>< / semantics>< / com:Math> vector, <com:Math><semantics>(h→l,k⊛g→l,k)<annotation encoding="application / x-tex">(\vec{h}_{l,k} \circledast \vec{g}_{l,k})< / annotation>< / semantics>< / com:Math> 1215, 1216 for <com:Math><semantics>l≤ℵk<annotation encoding="application / x-tex">l \leq \aleph_k< / annotation>< / semantics>< / com:Math>, formed as a circular convolution (denoted by '*) 1202, 1205) between a zero mean pseudo-random (PR) vector, <com:Math><semantics>g→l,k∈ℂNl×1<annotation encoding="application / x-tex">\vec{g}_{l,k} \in \mathbb{C}^{N_l \times 1}< / annotation>< / semantics>< / com:Math> 1203, 1206 with a DOD = 0, and a vector pulse, <com:Math><semantics>h→l,k∈ℂNl×1<annotation encoding="application / x-tex">\vec{h}_{l,k} \in \mathbb{C}^{N_l \times 1}< / annotation>< / semantics>< / com:Math> 1201, 1204 with a DOD = 1; while <com:Math><semantics>b→k∈ℂNℵ+1×1<annotation encoding="application / x-tex">\vec{b}_k \in \mathbb{C}^{N_{\aleph+1}\times 1}< / annotation>< / semantics>< / com:Math> 1215 is a zero mean PR vector with a DOD = 0. The first <com:Math><semantics>ℵk−1<annotation encoding="application / x-tex">\aleph_k - 1< / annotation>< / semantics>< / com:Math> linear convolutions produce <com:Math><semantics>NS≜∑l=1ℵNl−ℵk+1<annotation encoding="application / x-tex">N_S \triangleq \sum_{l=1}^{\aleph} N_l - \aleph_k + 1< / annotation>< / semantics>< / com:Math> S-FAT DOF, while the last produces <com:Math><semantics>Nℐ≜Nℵ+1−1<annotation encoding="application / x-tex">N_{\mathcal{I}} \triangleq N_{\aleph+1} - 1< / annotation>< / semantics>< / com:Math> I-FAT DOF. Vector II: <com:Math><semantics>℘→k,0∈ℂN0×1<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,0} \in \mathbb{C}^{N_0 \times 1}< / annotation>< / semantics>< / com:Math> 1209 is a PR vector with a DOD = 0 selected s.t. <com:Math><semantics>{hMTF}k<annotation encoding="application / x-tex">\{h_{MTF}\}_k< / annotation>< / semantics>< / com:Math> complies with the FOSE 707 constraint of <com:Math><semantics>ℙMask(f)<annotation encoding="application / x-tex">\mathbb{P}_{Mask}(f)< / annotation>< / semantics>< / com:Math> 701, i.e. with a power level <com:Math><semantics>𝒮xk(t)(f)≤1p<annotation encoding="application / x-tex">\mathcal{S}_{x_k(t)}(f) \leq \frac{1}{p}< / annotation>< / semantics>< / com:Math> in the FOSE band 707. It is possible to generalize <com:Math><semantics>N0<annotation encoding="application / x-tex">N_0< / annotation>< / semantics>< / com:Math> so that it is not necessarily equal to N. For example, it is possible to select <com:Math><semantics>N0=0<annotation encoding="application / x-tex">N_0 = 0< / annotation>< / semantics>< / com:Math>, implying that <com:Math><semantics>℘→k,0<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,0}< / annotation>< / semantics>< / com:Math> 1209 is not included in (8), or equivalently <com:Math><semantics>{hB_MTF}k≜℘→k,ℵk<annotation encoding="application / x-tex">\{h_{B\_MTF}\}_k \triangleq \overrightarrow{\wp}_{k,\aleph_k}< / annotation>< / semantics>< / com:Math>. It is also possible to select <com:Math><semantics>N0>N<annotation encoding="application / x-tex">N_0 > N< / annotation>< / semantics>< / com:Math>. In this case, <com:Math><semantics>N0−N<annotation encoding="application / x-tex">N_0 - N< / annotation>< / semantics>< / com:Math> zeros must be appended to <com:Math><semantics>℘→k,ℵk<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,\aleph_k}< / annotation>< / semantics>< / com:Math> 1208 in (9) in order for <com:Math><semantics>℘→k,ℵk<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,\aleph_k}< / annotation>< / semantics>< / com:Math> 1208 and <com:Math><semantics>{hB_MTF}k<annotation encoding="application / x-tex">\{h_{B\_MTF}\}_k< / annotation>< / semantics>< / com:Math> to have a total length of <com:Math><semantics>N0<annotation encoding="application / x-tex">N_0< / annotation>< / semantics>< / com:Math>. The reasoning behind separating <com:Math><semantics>{hB_MTF}k<annotation encoding="application / x-tex">\{h_{B\_MTF}\}_k< / annotation>< / semantics>< / com:Math> in (8) into two vectors, <com:Math><semantics>℘→k,ℵk<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,\aleph_k}< / annotation>< / semantics>< / com:Math> 1208 and <com:Math><semantics>℘→k,0<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,0}< / annotation>< / semantics>< / com:Math> 1209, is that it is difficult to simultaneously comply with the BW constraint, i.e. with <com:Math><semantics>WTL≤Wm<annotation encoding="application / x-tex">W_{TL} \leq W_m< / annotation>< / semantics>< / com:Math> 705; and the FOSE 707 constraint, i.e. with <com:Math><semantics>Sxk(t)(f)≤1R<annotation encoding="application / x-tex">S_{x_k(t)}(f) \leq \frac{1}{R}< / annotation>< / semantics>< / com:Math> using a single vector with a single DOD. By taking advantage of DOD Property I, summing <com:Math><semantics>℘→k,ℵk<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,\aleph_k}< / annotation>< / semantics>< / com:Math> 1208 and <com:Math><semantics>℘→k,0<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,0}< / annotation>< / semantics>< / com:Math> 1209 results in <com:Math><semantics>{hMTF}k<annotation encoding="application / x-tex">\{h_{MTF}\}_k< / annotation>< / semantics>< / com:Math> having a DOD <com:Math><semantics>ℵk<annotation encoding="application / x-tex">\aleph_k< / annotation>< / semantics>< / com:Math>, since <com:Math><semantics>℘→k,0<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,0}< / annotation>< / semantics>< / com:Math> 1209 has a DOD 0. The reasoning behind using <com:Math><semantics>ℵk−1<annotation encoding="application / x-tex">\aleph_k - 1< / annotation>< / semantics>< / com:Math> circular convolutions 1202, 1205 in <com:Math><semantics>℘→k,ℵk<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,\aleph_k}< / annotation>< / semantics>< / com:Math> 1208 in (8) is that it is difficult to use a single vector with a single DOD while achieving the following 2 requirements: (1) the entries of <com:Math><semantics>{hB_MTF}k<annotation encoding="application / x-tex">\{h_{B\_MTF}\}_k< / annotation>< / semantics>< / com:Math> are zero mean RVs; while (2) <com:Math><semantics>{hMTF}k<annotation encoding="application / x-tex">\{h_{MTF}\}_k< / annotation>< / semantics>< / com:Math> complies with the BW constraint that <com:Math><semantics>WTL≤Wm<annotation encoding="application / x-tex">W_{TL} \leq W_m< / annotation>< / semantics>< / com:Math> 705. By taking advantage of DOD Property II, circularly convolving <com:Math><semantics>h→l,k<annotation encoding="application / x-tex">\vec{h}_{l,k}< / annotation>< / semantics>< / com:Math> 1201, 1204 with <com:Math><semantics>g→l,k<annotation encoding="application / x-tex">\vec{g}_{l,k}< / annotation>< / semantics>< / com:Math> 1203, 1206, produces a vector with a DOD = 1 since the DOD for <com:Math><semantics>g→l,k<annotation encoding="application / x-tex">\vec{g}_{l,k}< / annotation>< / semantics>< / com:Math> 1203, 1206 is 0, implying that <com:Math><semantics>℘→k,ℵk<annotation encoding="application / x-tex">\vec{\wp}_{k,\aleph_k}< / annotation>< / semantics>< / com:Math> 1208 has a DOD <com:Math><semantics>ℵk<annotation encoding="application / x-tex">\aleph_k< / annotation>< / semantics>< / com:Math>. The pulse, <com:Math><semantics>℘→k,ℵk<annotation encoding="application / x-tex">\vec{\wp}_{k,\aleph_k}< / annotation>< / semantics>< / com:Math> 1208, is made to comply with <com:Math><semantics>WTL≤Wm<annotation encoding="application / x-tex">W_{TL} \leq W_m< / annotation>< / semantics>< / com:Math> 705 by properly selecting <com:Math><semantics>ℵk<annotation encoding="application / x-tex">\aleph_k< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>r<annotation encoding="application / x-tex">r< / annotation>< / semantics>< / com:Math>. Theorem IV: The MTF channel corresponding to <com:Math><semantics>{hB_MTF}k<annotation encoding="application / x-tex">\{h_{B\_MTF}\}_k< / annotation>< / semantics>< / com:Math> in (8) with <com:Math><semantics>Nℐ=0<annotation encoding="application / x-tex">N_{\mathcal{I}} = 0< / annotation>< / semantics>< / com:Math> under Constraints 1-3, has a capacity, <com:Math><semantics>𝒞MTF<annotation encoding="application / x-tex">\mathcal{C}_{MTF}< / annotation>< / semantics>< / com:Math>, identical to <com:Math><semantics>𝒞TL<annotation encoding="application / x-tex">\mathcal{C}_{TL}< / annotation>< / semantics>< / com:Math> in (7) except <com:Math><semantics>pk<annotation encoding="application / x-tex">p_k< / annotation>< / semantics>< / com:Math> is proportional to [Image disponible dans le document PDF, Image available in the PDF document] (10) where <com:Math><semantics>cK,N,d,ℵ<annotation encoding="application / x-tex">c_{K,N,d,\aleph}< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>cN0,0<annotation encoding="application / x-tex">c_{N_0,0}< / annotation>< / semantics>< / com:Math> are the coefficients of proportionality corresponding to <com:Math><semantics>℘→k,ℵ<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,\aleph}< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>℘→k,0<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,0}< / annotation>< / semantics>< / com:Math> in (8) respectively, with <com:Math><semantics>λ≜𝔭 / 7<annotation encoding="application / x-tex">\lambda \triangleq \mathfrak{p} / 7< / annotation>< / semantics>< / com:Math>, using an MMSE-SIC detector 110, 210, 310 at Rx 108, 208, 308. Importance of Theorem IV: Based on (10), <com:Math><semantics>CMTF<annotation encoding="application / x-tex">C_{MTF}< / annotation>< / semantics>< / com:Math> consists of two medium SNR regions 1006, 1007, as shown in FIG. 10b. As N increases, doubling <com:Math><semantics>CMTF<annotation encoding="application / x-tex">C_{MTF}< / annotation>< / semantics>< / com:Math> initially requires increasing the SNR by a fixed multiple of <com:Math><semantics>22ℵk<annotation encoding="application / x-tex">2^{2\aleph_k}< / annotation>< / semantics>< / com:Math>, which is the first medium SNR region 1006, referred to as "1st Med. SNR" in FIG. 10b. Then, later on it requires increasing the average received SNR by a fixed multiple of 2, which is the second medium SNR region 1007, referred to as "2<com:Sup>nd< / com:Sup> Med. SNR" in Fig. 10a. FIG. 10a compares <com:Math><semantics>𝒞MTF / 𝒲m<annotation encoding="application / x-tex">\mathcal{C}_{MTF} / \mathcal{W}_m< / annotation>< / semantics>< / com:Math> based on (10) with <com:Math><semantics>𝒞BL / 𝒲m<annotation encoding="application / x-tex">\mathcal{C}_{BL} / \mathcal{W}_m< / annotation>< / semantics>< / com:Math> in (6), where <com:Math><semantics>r=rBL=<annotation encoding="application / x-tex">r = r_{BL} =< / annotation>< / semantics>< / com:Math> 1, <com:Math><semantics>ℵ=1<annotation encoding="application / x-tex">\aleph = 1< / annotation>< / semantics>< / com:Math>, <com:Math><semantics>d=K=1<annotation encoding="application / x-tex">d = K = 1< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>℘→k,ℵ<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,\aleph}< / annotation>< / semantics>< / com:Math> in (8) is a rectangular pulse with <com:Math><semantics>℘→k,0<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,0}< / annotation>< / semantics>< / com:Math> in (8) <com:Math><semantics>30dBr<annotation encoding="application / x-tex">30dBr< / annotation>< / semantics>< / com:Math> below <com:Math><semantics>℘→k,ℵ<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,\aleph}< / annotation>< / semantics>< / com:Math>. In Fig 10b, <com:Math><semantics>𝒞MTF / 𝒲m<annotation encoding="application / x-tex">\mathcal{C}_{MTF} / \mathcal{W}_m< / annotation>< / semantics>< / com:Math> without Constraint 3 is illustrated as a number of curves (with "." markers), each curve corresponding to a value of <com:Math><semantics>Nmin<annotation encoding="application / x-tex">N_{min}< / annotation>< / semantics>< / com:Math>. <com:Math><semantics>CMTF / Wm<annotation encoding="application / x-tex">C_{MTF} / W_m< / annotation>< / semantics>< / com:Math> with Constraint 3 is illustrated as a number of points (with "square" markers), each point corresponding to a value of <com:Math><semantics>Nmin<annotation encoding="application / x-tex">N_{min}< / annotation>< / semantics>< / com:Math>. FIG. 10b confirms that indeed Constraint 3 maximizes <com:Math><semantics>𝒞MTF / 𝒲m<annotation encoding="application / x-tex">\mathcal{C}_{MTF} / \mathcal{W}_m< / annotation>< / semantics>< / com:Math> in both medium SNR regions 1006, 1007. Under certain conditions, the following asymptotic limits can be reached: a) When <com:Math><semantics>N0λR≪1<annotation encoding="application / x-tex">\frac{N_0}{\lambda R} \ll 1< / annotation>< / semantics>< / com:Math>, Theorem IV reduces to Theorem III. b) When <com:Math><semantics>N0λℛ≫1<annotation encoding="application / x-tex">\frac{N_0}{\lambda \mathcal{R}} \gg 1< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>cN0,0rN0λℛN≫1cK,N,d,ℵ1λrK2ℵk−1Nd2ℵk<annotation encoding="application / x-tex">c_{N_0,0} \frac{rN_0}{\lambda \mathcal{R}N} \gg \frac{1}{c_{K,N,d,\aleph} \frac{1}{\lambda r} K^{2\aleph_k - 1} N_d^{2\aleph_k}}< / annotation>< / semantics>< / com:Math>, we have [Image disponible dans le document PDF, Image available in the PDF document] (11) This limit applies to the case when <com:Math><semantics>ℙMask(f)<annotation encoding="application / x-tex">\mathbb{P}_{Mask}(f)< / annotation>< / semantics>< / com:Math> 701 corresponds to the IEEE 802.11 WLAN mask, denoted as <com:Math><semantics>ℙWiFi(f)<annotation encoding="application / x-tex">\mathbb{P}_{WiFi}(f)< / annotation>< / semantics>< / com:Math>. c) When <com:Math><semantics>N0λℛ≜q≪1<annotation encoding="application / x-tex">\frac{N_0}{\lambda \mathcal{R}} \triangleq q \ll 1< / annotation>< / semantics>< / com:Math> with <com:Math><semantics>q<annotation encoding="application / x-tex">q< / annotation>< / semantics>< / com:Math> a constant and <com:Math><semantics>cN0,0rN0λℛN≫1cK,N,d,ℵk1xK2ℵk−1Nd2ℵk<annotation encoding="application / x-tex">c_{N_0,0} \frac{rN_0}{\lambda \mathcal{R}N} \gg \frac{1}{c_{K,N,d,\aleph_k} \frac{1}{x} K^{2\aleph_k - 1} N_d^{2\aleph_k}}< / annotation>< / semantics>< / com:Math>, we have [Image disponible dans le document PDF, Image available in the PDF document] (12) This limit applies to the case when <com:Math><semantics>ℙMask(f)<annotation encoding="application / x-tex">\mathbb{P}_{Mask}(f)< / annotation>< / semantics>< / com:Math> 701 corresponds to the 3GPP LTE (E-UTRA) mask, denoted as <com:Math><semantics>ℙLTE(f)<annotation encoding="application / x-tex">\mathbb{P}_{LTE}(f)< / annotation>< / semantics>< / com:Math>. 5.3 MTF ARCHITECTURE Section 5.3.1 introduces the constraints that are generally imposed on communication systems such as standard-imposed spectral masks 701, as well as the effects of fading and interference across the communication channel 106, 306, 406. Section 5.3.2 proposes several MTF designs based on the constraints introduced in Section 5.3.1, while Section 5.3.3 introduces an architecture that is suitable for allowing various MTF systems to communicate with each other when co-located while using the same band. 5.3.1 Design Constraints: First, we select 2 important <com:Math><semantics>ℙMask(f)<annotation encoding="application / x-tex">\mathbb{P}_{Mask}(f)< / annotation>< / semantics>< / com:Math> 701, namely <com:Math><semantics>ℙWiFi(f)<annotation encoding="application / x-tex">\mathbb{P}_{WiFi}(f)< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>ℙLTE(f)<annotation encoding="application / x-tex">\mathbb{P}_{LTE}(f)< / annotation>< / semantics>< / com:Math>. Then, we model the communication channel 106, 306, 406 and examine its effects on the MTF architecture including the types of interference and restricted bands across such a channel. Selection of <com:Math><semantics>ℙMask(f)<annotation encoding="application / x-tex">\mathbb{P}_{Mask}(f)< / annotation>< / semantics>< / com:Math> 107: In order to include Constraint 2 in Design Steps I-III 1302, 1304, 1306, and in order to derive a fair comparison with some of the existing systems, we define <com:Math><semantics>ℙWiFi(f)<annotation encoding="application / x-tex">\mathbb{P}_{WiFi}(f)< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>ℙLTE(f)<annotation encoding="application / x-tex">\mathbb{P}_{LTE}(f)< / annotation>< / semantics>< / com:Math>: a) The 3GPP LTE (E-UTRA) mask, <com:Math><semantics>ℙLTE(f)<annotation encoding="application / x-tex">\mathbb{P}_{LTE}(f)< / annotation>< / semantics>< / com:Math>, is defined for a 1.4, 3, 5, 10, 15 and 20MHz BW, as having <com:Math><semantics>≤1%<annotation encoding="application / x-tex">\leq 1\%< / annotation>< / semantics>< / com:Math> OOBE BW, or equivalently, <com:Math><semantics>Wm<annotation encoding="application / x-tex">W_m< / annotation>< / semantics>< / com:Math> must contain <com:Math><semantics>≥99%<annotation encoding="application / x-tex">\geq 99\%< / annotation>< / semantics>< / com:Math> of the total integrated mean power in <com:Math><semantics>xk(t)<annotation encoding="application / x-tex">x_k(t)< / annotation>< / semantics>< / com:Math> 105, 205, 305 <com:Math><semantics>∀k<annotation encoding="application / x-tex">\forall k< / annotation>< / semantics>< / com:Math>. b) The IEEE 802.11 WLAN mask for a 20MHz BW is: [Image disponible dans le document PDF, Image available in the PDF document] In <com:Math><semantics>ℙWiFi(f)<annotation encoding="application / x-tex">\mathbb{P}_{WiFi}(f)< / annotation>< / semantics>< / com:Math>, the first frequency band, <com:Math><semantics>|f|≤9MHz<annotation encoding="application / x-tex">|f| \leq 9MHz< / annotation>< / semantics>< / com:Math>, corresponds to the occupied band 705 with a bandwidth, <com:Math><semantics>Wm=18MHz<annotation encoding="application / x-tex">W_m = 18MHz< / annotation>< / semantics>< / com:Math>. The middle three frequency bands correspond to OOBE 706 with a bandwidth, <com:Math><semantics>WOOBE=42MHz<annotation encoding="application / x-tex">W_{OOBE} = 42MHz< / annotation>< / semantics>< / com:Math>. The last frequency band, <com:Math><semantics>|f|≥30MHz<annotation encoding="application / x-tex">|f| \ge 30MHz< / annotation>< / semantics>< / com:Math>, correspond to the FOSE band 707 with an infinite bandwidth and a power level <com:Math><semantics>1 / ℛ=<annotation encoding="application / x-tex">1 / \mathcal{R} =< / annotation>< / semantics>< / com:Math> <com:Math><semantics>−40dBr<annotation encoding="application / x-tex">-40dBr< / annotation>< / semantics>< / com:Math>. Modeling of the Communication Channel 106, 306, 406: When <com:Math><semantics>fs≜1 / Ts<annotation encoding="application / x-tex">f_s \triangleq 1 / T_s< / annotation>< / semantics>< / com:Math> is <com:Math><semantics>≥8MHz<annotation encoding="application / x-tex">\geq 8MHz< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>NTs≤1ms<annotation encoding="application / x-tex">NT_s \leq 1ms< / annotation>< / semantics>< / com:Math>, the channel 106, 306, 406 can be modeled as a frequency-selective (FS) slowly fading channel affected by a frequency-dependent path loss (PL) modeled after Friis free-space PL (FSPL) model. Mathematically, such a channel can be modeled as Linear Time-Invariant (LTI) and characterized using a discrete-time random impulse response, <com:Math><semantics>h→Ch<annotation encoding="application / x-tex">\vec{h}_{Ch}< / annotation>< / semantics>< / com:Math>, of finite length, <com:Math><semantics>δN<annotation encoding="application / x-tex">\delta N< / annotation>< / semantics>< / com:Math>, referred to as the discrete delay spread of the channel. The fading can be modeled either as Rayleigh for a non-LOS channel or as Rician with a strong LOS component for a LOS channel. Effects of the Selected Channel model: 1) Mathematically, the main effect of the frequency-selective channel is to linearly convolve each column, <com:Math><semantics>{hMTF}k<annotation encoding="application / x-tex">\{h_{MTF}\}_k< / annotation>< / semantics>< / com:Math>, in <com:Math><semantics>hMTF<annotation encoding="application / x-tex">h_{MTF}< / annotation>< / semantics>< / com:Math> 1104 with <com:Math><semantics>h→Ch<annotation encoding="application / x-tex">\vec{h}_{Ch}< / annotation>< / semantics>< / com:Math>. The outcome of such a convolution is a new MTF matrix, <com:Math><semantics>𝒉MTF,Ch∈ℂℳ×Q<annotation encoding="application / x-tex">\mathbf{h}_{MTF,Ch} \in \mathbb{C}^{\mathcal{M} \times Q}< / annotation>< / semantics>< / com:Math>, defined as [Image disponible dans le document PDF, Image available in the PDF document] (13) where <com:Math><semantics>h→MTF,Ch∈ℂℳ×K<annotation encoding="application / x-tex">\vec{h}_{MTF,Ch} \in \mathbb{C}^{\mathcal{M} \times K}< / annotation>< / semantics>< / com:Math> has <com:Math><semantics>hBMTF<annotation encoding="application / x-tex">h_{BMTF}< / annotation>< / semantics>< / com:Math> replaced by <com:Math><semantics>hBMTF,Ch∈ℂ𝒩×K<annotation encoding="application / x-tex">h_{BMTF,Ch} \in \mathbb{C}^{\mathcal{N} \times K}< / annotation>< / semantics>< / com:Math> with <com:Math><semantics>M<annotation encoding="application / x-tex">M< / annotation>< / semantics>< / com:Math> replaced by <com:Math><semantics>ℳ≜<annotation encoding="application / x-tex">\mathcal{M} \triangleq< / annotation>< / semantics>< / com:Math> <com:Math><semantics>𝒩+d(L−1)<annotation encoding="application / x-tex">\mathcal{N} + d(L-1)< / annotation>< / semantics>< / com:Math>, N replaced by <com:Math><semantics>𝒩≜N+δN−1<annotation encoding="application / x-tex">\mathcal{N} \triangleq N + \delta N - 1< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>No<annotation encoding="application / x-tex">N_o< / annotation>< / semantics>< / com:Math> replaced by <com:Math><semantics>𝒩o≜No+δN−1<annotation encoding="application / x-tex">\mathcal{N}_o \triangleq N_o + \delta N - 1< / annotation>< / semantics>< / com:Math>. The increase in N and in <com:Math><semantics>No<annotation encoding="application / x-tex">N_o< / annotation>< / semantics>< / com:Math> by <com:Math><semantics>δN−1<annotation encoding="application / x-tex">\delta N - 1< / annotation>< / semantics>< / com:Math> is equivalent to an increase in the number, <com:Math><semantics>Nℐ<annotation encoding="application / x-tex">N_{\mathcal{I}}< / annotation>< / semantics>< / com:Math>, of I-FAT DOF in the MTF system by <com:Math><semantics>δN<annotation encoding="application / x-tex">\delta N< / annotation>< / semantics>< / com:Math>. Based on DOD Property II, the linear convolution between <com:Math><semantics>h→Ch<annotation encoding="application / x-tex">\vec{h}_{Ch}< / annotation>< / semantics>< / com:Math> and the <com:Math><semantics>kth<annotation encoding="application / x-tex">k^{th}< / annotation>< / semantics>< / com:Math> column, <com:Math><semantics>{hMTF}k<annotation encoding="application / x-tex">\{h_{MTF}\}_k< / annotation>< / semantics>< / com:Math>, in <com:Math><semantics>hMTF<annotation encoding="application / x-tex">h_{MTF}< / annotation>< / semantics>< / com:Math> 1104 implies that the resulting DOD is equal to the sum between the original DOD, <com:Math><semantics>ℵk<annotation encoding="application / x-tex">\aleph_k< / annotation>< / semantics>< / com:Math>, and the DOD, <com:Math><semantics>ℵCh<annotation encoding="application / x-tex">\aleph_{Ch}< / annotation>< / semantics>< / com:Math>, of the communication channel. Based on the adopted frequency-selective fading model of the communication channel 106, 306, 406, <com:Math><semantics>ℵCh=0<annotation encoding="application / x-tex">\aleph_{Ch} = 0< / annotation>< / semantics>< / com:Math>. In other words, the resulting DOD of <com:Math><semantics>hMTF,Ch<annotation encoding="application / x-tex">h_{MTF,Ch}< / annotation>< / semantics>< / com:Math>, is equal to the original DOD of <com:Math><semantics>hMTF<annotation encoding="application / x-tex">h_{MTF}< / annotation>< / semantics>< / com:Math> 1104 when the communication channel 106, 306, 406 is FS. 2) The communication channel 106, 306, 406 can be equivalently characterized in the frequency domain by the DTFT of <com:Math><semantics>h→Ch<annotation encoding="application / x-tex">\vec{h}_{Ch}< / annotation>< / semantics>< / com:Math>, also referred to as its Transfer Function (TF), <com:Math><semantics>HCh(Ω)<annotation encoding="application / x-tex">H_{Ch}(\Omega)< / annotation>< / semantics>< / com:Math>. This implies that the following continuous-frequency product: [Image disponible dans le document PDF, Image available in the PDF document] (14) can replace the linear discrete-time convolution between <com:Math><semantics>{hMTF}k<annotation encoding="application / x-tex">\{h_{MTF}\}_k< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>h→Ch<annotation encoding="application / x-tex">\vec{h}_{Ch}< / annotation>< / semantics>< / com:Math> where <com:Math><semantics>HMTFk,Ch(Ω)<annotation encoding="application / x-tex">H_{MTF_k,Ch}(\Omega)< / annotation>< / semantics>< / com:Math> is the DTFT of the <com:Math><semantics>kth<annotation encoding="application / x-tex">k^{th}< / annotation>< / semantics>< / com:Math> column, <com:Math><semantics>{𝒉MTF,Ch}k<annotation encoding="application / x-tex">\{\boldsymbol{h}_{MTF,Ch}\}_k< / annotation>< / semantics>< / com:Math>, of <com:Math><semantics>𝒉MTF,Ch<annotation encoding="application / x-tex">\boldsymbol{h}_{MTF,Ch}< / annotation>< / semantics>< / com:Math>. 3) Friis FSPL model is based on <com:Math><semantics>E{|HCh(Ω)|2}<annotation encoding="application / x-tex">E\{|H_{Ch}(\Omega)|^2\}< / annotation>< / semantics>< / com:Math> being inversely proportional to <com:Math><semantics>|Ω|2<annotation encoding="application / x-tex">|\Omega|^2< / annotation>< / semantics>< / com:Math>, i.e. [Image disponible dans le document PDF, Image available in the PDF document] (15) where <com:Math><semantics>E{.}<annotation encoding="application / x-tex">E\{.\}< / annotation>< / semantics>< / com:Math> denotes expectation w.r.t. <com:Math><semantics>HCh(Ω)<annotation encoding="application / x-tex">H_{Ch}(\Omega)< / annotation>< / semantics>< / com:Math> at <com:Math><semantics>Ω<annotation encoding="application / x-tex">\Omega< / annotation>< / semantics>< / com:Math>, assuming it is ergodic. Based on (15), it is possible to see that the FSPL has a DOD equal to 1. In other words, the effect of the FSPL is to increase the original DOD, <com:Math><semantics>ℵk<annotation encoding="application / x-tex">\aleph_k< / annotation>< / semantics>< / com:Math>, of <com:Math><semantics>{hMTF}k<annotation encoding="application / x-tex">\{h_{MTF}\}_k< / annotation>< / semantics>< / com:Math> by 1 if the carrier frequency, <com:Math><semantics>fc=0<annotation encoding="application / x-tex">f_c = 0< / annotation>< / semantics>< / com:Math>, otherwise, the effect of the FSPL on <com:Math><semantics>ℵk<annotation encoding="application / x-tex">\aleph_k< / annotation>< / semantics>< / com:Math> depends on <com:Math><semantics>fc<annotation encoding="application / x-tex">f_c< / annotation>< / semantics>< / com:Math>. Based on all effects of the communication channel 106, 306, 406, Theorem III is still valid after replacing <com:Math><semantics>HMTFk(Ω)<annotation encoding="application / x-tex">H_{MTF_k}(\Omega)< / annotation>< / semantics>< / com:Math> by <com:Math><semantics>HMTFk,Ch(Ω)<annotation encoding="application / x-tex">H_{MTF_k,Ch}(\Omega)< / annotation>< / semantics>< / com:Math>, M by <com:Math><semantics>ℳ<annotation encoding="application / x-tex">\mathcal{M}< / annotation>< / semantics>< / com:Math>, and after re-evaluating <com:Math><semantics>ℵk<annotation encoding="application / x-tex">\aleph_k< / annotation>< / semantics>< / com:Math> based on <com:Math><semantics>fc<annotation encoding="application / x-tex">f_c< / annotation>< / semantics>< / com:Math>. In order to preserve the original DOD, <com:Math><semantics>ℵk<annotation encoding="application / x-tex">\aleph_k< / annotation>< / semantics>< / com:Math>, of <com:Math><semantics>hMTF<annotation encoding="application / x-tex">h_{MTF}< / annotation>< / semantics>< / com:Math> 1104 a pre-channel filter 500 is required at Tx 104, 204, 304, which is discussed in Section 5.3.2. Modeling of Interference: Two types of interference exist across a communication network: (a) Narrow-Band Interference (NBI), defined as having a width <com:Math><semantics>≤125MHz<annotation encoding="application / x-tex">\leq 125MHz< / annotation>< / semantics>< / com:Math>; and (b) Wide-Band Interference (WBI), defined as having a width <com:Math><semantics>>125MHz<annotation encoding="application / x-tex">> 125MHz< / annotation>< / semantics>< / com:Math>. NBI encompasses transmissions from existing systems such as LTE and Wi-Fi systems, and from other MTF systems due to the presence of <com:Math><semantics>℘→k,ℵk<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,\aleph_k}< / annotation>< / semantics>< / com:Math> 1208, while WBI encompasses transmissions from Ultra-Wide Band (UWB) systems and from other MTF systems due to the presence of <com:Math><semantics>℘→k,0<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,0}< / annotation>< / semantics>< / com:Math> 1209. Several studies have indicated low utilization of the frequency bands at frequencies > 2GHz as seen in Table I, which displays the average duty cycle versus frequency range <com:Math><semantics>≤7,075MHz<annotation encoding="application / x-tex">\leq 7,075MHz< / annotation>< / semantics>< / com:Math> based on results in in an urban environment. Table I is consistent with several other studies of urban centers across North America and Europe. All studies indicate an exponential decline in utilization directly proportional to frequency, <com:Math><semantics>f<annotation encoding="application / x-tex">f< / annotation>< / semantics>< / com:Math>. We refer to frequency ranges with known Heavy Utilization as <com:Math><semantics>ℬHU<annotation encoding="application / x-tex">\mathcal{B}_{HU}< / annotation>< / semantics>< / com:Math>. [Image disponible dans le document PDF, Image available in the PDF document] Table I Average utilization duty cycle across a frequency range. Restricted Bands, <com:Math><semantics>ℬRR<annotation encoding="application / x-tex">\mathcal{B}_{RR}< / annotation>< / semantics>< / com:Math>: Further to having to contend with both NBI and WBI, some bands, referred to as <com:Math><semantics>ℬRB<annotation encoding="application / x-tex">\mathcal{B}_{RB}< / annotation>< / semantics>< / com:Math>, have been deemed restricted by the regulatory bodies (47 CFR 15.205). 5.3.2 Pulse and Filter Design: Based on the knowledge of the statistics of the communication channel 106, 306, 406 including its model, the types of interference across it and the existence of <com:Math><semantics>ℬRB<annotation encoding="application / x-tex">\mathcal{B}_{RB}< / annotation>< / semantics>< / com:Math>, the disclosure designs pulses such as <com:Math><semantics>h→l,k<annotation encoding="application / x-tex">\vec{h}_{l,k}< / annotation>< / semantics>< / com:Math> 1201, 1204, <com:Math><semantics>℘→k,0<annotation encoding="application / x-tex">\vec{\wp}_{k,0}< / annotation>< / semantics>< / com:Math> 1209, <com:Math><semantics>g→l,k<annotation encoding="application / x-tex">\vec{g}_{l,k}< / annotation>< / semantics>< / com:Math> 1203, 1206 as well as filters such as a Pre- channel filter 500 at Tx 104, 204, 304, and a Post-channel filter 615 at Rx 108, 208, 308 with the goal of optimizing <com:Math><semantics>CMTF<annotation encoding="application / x-tex">C_{MTF}< / annotation>< / semantics>< / com:Math> subject to Constraints 1-3. Design of <com:Math><semantics>h→l,k∈ℂNl×1<annotation encoding="application / x-tex">\vec{h}_{l,k} \in \mathbb{C}^{N_l \times 1}< / annotation>< / semantics>< / com:Math> 1201, 1204: A fundamental design for <com:Math><semantics>h→l,k<annotation encoding="application / x-tex">\vec{h}_{l,k}< / annotation>< / semantics>< / com:Math> 1201, 1204 in (9) is a rectangular pulse. Even though it has all its zeros on the unit circle, it is possible to move its zeros away from the unit circle by shifting it in frequency by <com:Math><semantics>1 / (2NlTs)<annotation encoding="application / x-tex">1 / (2N_lT_s)< / annotation>< / semantics>< / com:Math>. This shift forms a complex pulse, <com:Math><semantics>h→REC,l,k<annotation encoding="application / x-tex">\vec{h}_{REC,l,k}< / annotation>< / semantics>< / com:Math> 1212, 1213, with a real part in the shape of one lobe of a cosine wave and an imaginary part in the shape of a negative lobe of a sine wave. When all <com:Math><semantics>h→l,k∀l<annotation encoding="application / x-tex">\vec{h}_{l,k} \forall l< / annotation>< / semantics>< / com:Math> 1201, 1204 are selected as <com:Math><semantics>h→REC,l,k<annotation encoding="application / x-tex">\vec{h}_{REC,l,k}< / annotation>< / semantics>< / com:Math> 1212, 1213, <com:Math><semantics>℘→k,ℵk<annotation encoding="application / x-tex">\vec{\wp}_{k,\aleph_k}< / annotation>< / semantics>< / com:Math> 1208 in (9) is denoted as <com:Math><semantics>℘→REC,ℵk<annotation encoding="application / x-tex">\vec{\wp}_{REC,\aleph_k}< / annotation>< / semantics>< / com:Math> 1214. When <com:Math><semantics>℘→k,ℵk<annotation encoding="application / x-tex">\vec{\wp}_{k,\aleph_k}< / annotation>< / semantics>< / com:Math> 1208 in (9) is selected as <com:Math><semantics>℘→REC,1<annotation encoding="application / x-tex">\overrightarrow{\wp}_{REC,1}< / annotation>< / semantics>< / com:Math>, <com:Math><semantics>λ≜𝔭 / 7<annotation encoding="application / x-tex">\lambda \triangleq \mathfrak{p} / 7< / annotation>< / semantics>< / com:Math> in Theorem III is asymptotically equal to <com:Math><semantics>π<annotation encoding="application / x-tex">\pi< / annotation>< / semantics>< / com:Math>. In this case, the amplitude of <com:Math><semantics>℘→REC.1<annotation encoding="application / x-tex">\overrightarrow{\wp}_{REC.1}< / annotation>< / semantics>< / com:Math> is selected to comply with Constraint 1. Design of <com:Math><semantics>℘→k,0∈ℂN0×1<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,0} \in \mathbb{C}^{N_0 \times 1}< / annotation>< / semantics>< / com:Math> 1209: A possible design of <com:Math><semantics>℘→k,0<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,0}< / annotation>< / semantics>< / com:Math> 1209 in (8) is: [Image disponible dans le document PDF, Image available in the PDF document] (16) where <com:Math><semantics>𝒟ℱ𝒯−1<annotation encoding="application / x-tex">\mathcal{DFT}^{-1}< / annotation>< / semantics>< / com:Math> denotes an inverse Discrete Fourier Transform (DFT) operation; and the phase, <com:Math><semantics>ϑk,i∈{0,2π}<annotation encoding="application / x-tex">\vartheta_{k,i} \in \{0,2\pi\}< / annotation>< / semantics>< / com:Math>, is chosen as PR with a uniform distribution across <com:Math><semantics>{0,2π}<annotation encoding="application / x-tex">\{0,2\pi\}< / annotation>< / semantics>< / com:Math> for <com:Math><semantics>1≤i≤N0<annotation encoding="application / x-tex">1 \le i \le N_0< / annotation>< / semantics>< / com:Math>. <com:Math><semantics>℘→k,0<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,0}< / annotation>< / semantics>< / com:Math> 1209 is also known as a frequency-based PR polyphase signature. Design of <com:Math><semantics>g→l,k∈ℂNl×1<annotation encoding="application / x-tex">\vec{g}_{l,k} \in \mathbb{C}^{N_l \times 1}< / annotation>< / semantics>< / com:Math> 1203, 1206: A possible design of <com:Math><semantics>g→l,k<annotation encoding="application / x-tex">\vec{g}_{l,k}< / annotation>< / semantics>< / com:Math> 1203, 1206 in (9) is [Image disponible dans le document PDF, Image available in the PDF document] (17) with the phase, <com:Math><semantics>θl,k,i∈{0,2π}<annotation encoding="application / x-tex">\theta_{l,k,i} \in \{0,2\pi\}< / annotation>< / semantics>< / com:Math>, chosen as PR with a uniform distribution across <com:Math><semantics>{0,2π}<annotation encoding="application / x-tex">\{0,2\pi\}< / annotation>< / semantics>< / com:Math> for <com:Math><semantics>1≤1<annotation encoding="application / x-tex">1 \le 1< / annotation>< / semantics>< / com:Math> <com:Math><semantics>i≤Nl<annotation encoding="application / x-tex">i \leq N_l< / annotation>< / semantics>< / com:Math>, similar to <com:Math><semantics>℘→k,0<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,0}< / annotation>< / semantics>< / com:Math> 1209 in (16), except that <com:Math><semantics>ϑl,k,1<annotation encoding="application / x-tex">\vartheta_{l,k,1}< / annotation>< / semantics>< / com:Math> must equal <com:Math><semantics>ϑl,k,Nl<annotation encoding="application / x-tex">\vartheta_{l,k,N_l}< / annotation>< / semantics>< / com:Math> in order to minimize the overhead factor, <com:Math><semantics>𝒩<annotation encoding="application / x-tex">\mathcal{N}< / annotation>< / semantics>< / com:Math>. Since the communication channel forces <com:Math><semantics>HMTFk(Ω)<annotation encoding="application / x-tex">H_{MTF_k}(\Omega)< / annotation>< / semantics>< / com:Math> in (14) to be multipled by <com:Math><semantics>HCh(Ω)<annotation encoding="application / x-tex">H_{Ch}(\Omega)< / annotation>< / semantics>< / com:Math>, the resulting product, <com:Math><semantics>HMTFk,Ch(Ω)<annotation encoding="application / x-tex">H_{MTF_k,Ch}(\Omega)< / annotation>< / semantics>< / com:Math>, forces <com:Math><semantics>g→l,k<annotation encoding="application / x-tex">\vec{g}_{l,k}< / annotation>< / semantics>< / com:Math> 1203, 1206 in (17) to be replaced by [Image disponible dans le document PDF, Image available in the PDF document] (18) when <com:Math><semantics>Nℐ=1<annotation encoding="application / x-tex">N_{\mathcal{I}} = 1< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>ℵk=1<annotation encoding="application / x-tex">\aleph_k = 1< / annotation>< / semantics>< / com:Math> where <com:Math><semantics>|al,k,1|,…,|al,k,𝒩l|<annotation encoding="application / x-tex">|a_{l,k,1}|, \dots, |a_{l,k,\mathcal{N}_l}|< / annotation>< / semantics>< / com:Math> are random amplitudes, which have either a Rician distribution with a strong LOS component across a LOS channel, or a Rayleigh distribution across a NLOS channel, and <com:Math><semantics>𝒩l≜Nl+δN−1<annotation encoding="application / x-tex">\mathcal{N}_l \triangleq N_l + \delta N - 1< / annotation>< / semantics>< / com:Math>. Selection of <com:Math><semantics>q<annotation encoding="application / x-tex">q< / annotation>< / semantics>< / com:Math> when <com:Math><semantics>ℙMask(f)≡ℙLTE(f)<annotation encoding="application / x-tex">\mathbb{P}_{Mask}(f) \equiv \mathbb{P}_{LTE}(f)< / annotation>< / semantics>< / com:Math>: The disclosure selects <com:Math><semantics>q<annotation encoding="application / x-tex">q< / annotation>< / semantics>< / com:Math> in (12) as 0.5% and allocates the remaining 0.5% to the OOBE 706 BW in <com:Math><semantics>℘→k,ℵk<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,\aleph_k}< / annotation>< / semantics>< / com:Math> 1208. Under Constraint 2, (12) can be re-written as <com:Math><semantics>lim⁡N0→∞pk=(N / r)dB+23dB<annotation encoding="application / x-tex">\lim_{N_0 \to \infty} p_k = (N / r)_{dB} + 23dB< / annotation>< / semantics>< / com:Math>. Design of Pre-channel filtering 500 at Tx 104, 204, 304: In order to comply with <com:Math><semantics>ℬRB<annotation encoding="application / x-tex">\mathcal{B}_{RB}< / annotation>< / semantics>< / com:Math> (47 CFR 15.205), and to prevent transmitting across <com:Math><semantics>ℬHU<annotation encoding="application / x-tex">\mathcal{B}_{HU}< / annotation>< / semantics>< / com:Math>, a pre-channel filter 500 is recommended at Tx 104, 204, 304. Furthermore, according to (15), the effect of the FSPL is to increase the DOD, <com:Math><semantics>ℵk<annotation encoding="application / x-tex">\aleph_k< / annotation>< / semantics>< / com:Math>, of <com:Math><semantics>hMTF<annotation encoding="application / x-tex">h_{MTF}< / annotation>< / semantics>< / com:Math> 1104 by 1 despite the fact that <com:Math><semantics>℘→k,0<annotation encoding="application / x-tex">\vec{\wp}_{k,0}< / annotation>< / semantics>< / com:Math> 1209 has been added in (8) to force the resulting DOD to asymptotically take the value 0. In order to address all 3 concerns, <com:Math><semantics>hMTF<annotation encoding="application / x-tex">h_{MTF}< / annotation>< / semantics>< / com:Math> 1104 is replaced by a pre-channel MTF matrix, <com:Math><semantics>hMTF𝒫<annotation encoding="application / x-tex">h_{MTF}^{\mathcal{P}}< / annotation>< / semantics>< / com:Math>, based on replacing in (8) <com:Math><semantics>℘→k,ℵk<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,\aleph_k}< / annotation>< / semantics>< / com:Math> 1208 by <com:Math><semantics>℘→k,ℵk𝒫<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,\aleph_k}^{\mathcal{P}}< / annotation>< / semantics>< / com:Math>, <com:Math><semantics>℘→k,0<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,0}< / annotation>< / semantics>< / com:Math> 1209 by <com:Math><semantics>℘→k,0𝒫<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,0}^{\mathcal{P}}< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>{𝒉MTF}k<annotation encoding="application / x-tex">\{\boldsymbol{h}_{MTF}\}_k< / annotation>< / semantics>< / com:Math> by <com:Math><semantics>{𝒉MTF𝒫}k≜℘→k,ℵ𝒫+℘→k,0𝒫<annotation encoding="application / x-tex">\{\boldsymbol{h}_{MTF}^{\mathcal{P}}\}_k \triangleq \overrightarrow{\wp}_{k,\aleph}^{\mathcal{P}} + \overrightarrow{\wp}_{k,0}^{\mathcal{P}}< / annotation>< / semantics>< / com:Math> with a DTFT <com:Math><semantics>HMTFk𝒫(Ω)<annotation encoding="application / x-tex">H_{MTF_k}^{\mathcal{P}}(\Omega)< / annotation>< / semantics>< / com:Math> pre- processed by the following 2 actions: 1. Pre-distort <com:Math><semantics>HMTFk𝒫(Ω)<annotation encoding="application / x-tex">H_{MTF_k}^{\mathcal{P}}(\Omega)< / annotation>< / semantics>< / com:Math> by <com:Math><semantics>|Ω|∀Ω∈[−π,0)∪(0,π]<annotation encoding="application / x-tex">|\Omega| \ \forall \Omega \in [-\pi, 0) \cup (0, \pi]< / annotation>< / semantics>< / com:Math> as [Image disponible dans le document PDF, Image available in the PDF document] (19) where <com:Math><semantics>ν<annotation encoding="application / x-tex">\nu< / annotation>< / semantics>< / com:Math> is selected to keep <com:Math><semantics>E{|HMTFk𝒫(Ω)|2}≤1𝒫∀Ω∈[−π,0)∪(0,π].<annotation encoding="application / x-tex">E\left\{\left|H_{MTF_{k}}^{\mathcal{P}}(\Omega)\right|^{2}\right\} \leq \frac{1}{\mathcal{P}} \,\forall \Omega \in [-\pi, 0) \cup (0, \pi].< / annotation>< / semantics>< / com:Math> 2. Force <com:Math><semantics>HMTFν𝒫(Ω)<annotation encoding="application / x-tex">H_{MTF_{\nu}}^{\mathcal{P}}(\Omega)< / annotation>< / semantics>< / com:Math> to contain a null at the excluded band: <com:Math><semantics>ℬex≜{{ℬRB∪ℬHU}∩ℬTL}∪ℬ¯TL<annotation encoding="application / x-tex">\mathcal{B}_{ex} \triangleq \{\{\mathcal{B}_{RB} \cup \mathcal{B}_{HU}\} \cap \mathcal{B}_{TL}\} \cup \overline{\mathcal{B}}_{TL}< / annotation>< / semantics>< / com:Math> with <com:Math><semantics>ℬHU≜[0,2GHz]<annotation encoding="application / x-tex">\mathcal{B}_{HU} \triangleq [0,2GHz]< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>ℬ¯TL<annotation encoding="application / x-tex">\overline{\mathcal{B}}_{TL}< / annotation>< / semantics>< / com:Math> the complement of <com:Math><semantics>ℬTL≜[fc−fs / 4,fc+fs / 4]<annotation encoding="application / x-tex">\mathcal{B}_{TL} \triangleq [f_c - f_s / 4, f_c + f_s / 4]< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>fc≥fs / 4<annotation encoding="application / x-tex">f_c \geq f_s / 4< / annotation>< / semantics>< / com:Math>. As a result of both actions, <com:Math><semantics>pk<annotation encoding="application / x-tex">p_k< / annotation>< / semantics>< / com:Math> in (10) is replaced by [Image disponible dans le document PDF, Image available in the PDF document] (20) and (11) is replaced by <com:Math><semantics>lim⁡N0→∞pk→(Nr)dR−5.16dB<annotation encoding="application / x-tex">\lim_{N_0 \to \infty} p_k \to \left(\frac{N}{r}\right)_{dR} - 5.16dB< / annotation>< / semantics>< / com:Math> since <com:Math><semantics>v=2(π2 / 6)<annotation encoding="application / x-tex">v = 2(\pi^2 / 6)< / annotation>< / semantics>< / com:Math> according to Basel problem where <com:Math><semantics>2(lim⁡N0→∞∑n=1N01n2)=2(π2 / 6)=5.16dB<annotation encoding="application / x-tex">2\left(\lim_{N_0\to\infty}\sum_{n=1}^{N_0}\frac{1}{n^2}\right)=2(\pi^2 / 6)=5.16dB< / annotation>< / semantics>< / com:Math> when the samples are real. On the other hand, (12) is replaced by <com:Math><semantics>lim⁡N0→∞pk=(Nr)dB−(q)dB−5.16dB<annotation encoding="application / x-tex">\lim_{N_0 \to \infty} p_k = \left(\frac{N}{r}\right)_{dB} - (q)_{dB} - 5.16dB< / annotation>< / semantics>< / com:Math>. For example, when <com:Math><semantics>q=0.5%<annotation encoding="application / x-tex">q = 0.5\%< / annotation>< / semantics>< / com:Math>, <com:Math><semantics>lim⁡N0→∞𝒫k=(Nr)dR+17.9dB.<annotation encoding="application / x-tex">\lim_{N_0 \to \infty} \mathcal{P}_k = \left(\frac{N}{r}\right)_{dR} + 17.9dB.< / annotation>< / semantics>< / com:Math> Design of Post-channel filtering 615 at Rx 108, 208, 308: Post-channel filtering 615 can be used at Rx 108, 208, 308 to reduce the effects of NBI across the communication channel. In this case, it must include an MTF excision filter, which consists of the following two steps: (a) Estimate the frequency range, <com:Math><semantics>ℬNBI<annotation encoding="application / x-tex">\mathcal{B}_{NBI}< / annotation>< / semantics>< / com:Math>, corresponding to NBI. A frequency, f, belongs to <com:Math><semantics>ℬNBI<annotation encoding="application / x-tex">\mathcal{B}_{NBI}< / annotation>< / semantics>< / com:Math> when <com:Math><semantics>Sy(t)(f)≥Sxk(t)(f)+STh<annotation encoding="application / x-tex">S_{y(t)}(f) \geq S_{x_k(t)}(f) + S_{Th}< / annotation>< / semantics>< / com:Math> where <com:Math><semantics>STh<annotation encoding="application / x-tex">S_{Th}< / annotation>< / semantics>< / com:Math> is a threshold selected to meet a certain optimization criterion for reducing NBI. (b) Excise the estimated NBI by forcing a null in the PSD, <com:Math><semantics>Sy(t)(f)<annotation encoding="application / x-tex">S_{y(t)}(f)< / annotation>< / semantics>< / com:Math>, of the continuous-time version, <com:Math><semantics>y(t)<annotation encoding="application / x-tex">y(t)< / annotation>< / semantics>< / com:Math>, of <com:Math><semantics>r→<annotation encoding="application / x-tex">\vec{r}< / annotation>< / semantics>< / com:Math>, at <com:Math><semantics>f∈ℬNBI<annotation encoding="application / x-tex">f \in \mathcal{B}_{NBI}< / annotation>< / semantics>< / com:Math>. Post-channel filtering should also include a null at <Formule mathématique disponible dans le document PDF, Math available in the PDF document>\mathcal{B}_{NBI} \cup \mathcal{B}_{ex} \triangleq \{\{\mathcal{B}_{NBI} \cup \mathcal{B}_{RB} \cup \mathcal{B}_{HU}\} \cap \mathcal{B}_{ex} \in \{\mathcal{B}_{NBI} \cup \mathcal{B}_{RB} \cup \mathcal{B}_{HU}\} \cap \mathcal{B}_{ex} \in \mathcal{B}_{ex} \in \mathcal{B}_{ex} \cup \mathcal{B}_{ex} \in \mathcal{B <com:Math><semantics>ℬTL<annotation encoding="application / x-tex">\mathcal{B}_{TL}< / annotation>< / semantics>< / com:Math> <com:Math><semantics>∪ℬ¯TL<annotation encoding="application / x-tex">\cup \overline{\mathcal{B}}_{TL}< / annotation>< / semantics>< / com:Math> in order to reduce the effect of noise and interference at Rx 108, 208, 308. Selection of Sampling Type and frequency, <com:Math><semantics>fs<annotation encoding="application / x-tex">f_s< / annotation>< / semantics>< / com:Math>: There are 3 types of sampling available in communication systems: baseband sampling, IF sampling and RF sampling. RF sampling is recommended when <com:Math><semantics>fs≥4fc<annotation encoding="application / x-tex">f_s \ge 4f_c< / annotation>< / semantics>< / com:Math>, since it does not require any up-conversion / down-conversion stages as shown in FIGs. 6a and 6b. In this case, the Carrier Frequency, <com:Math><semantics>fc<annotation encoding="application / x-tex">f_c< / annotation>< / semantics>< / com:Math>, is selected <com:Math><semantics>≤fs / 4<annotation encoding="application / x-tex">\leq f_s / 4< / annotation>< / semantics>< / com:Math>. On the other hand, IF sampling is recommended over baseband sampling when <com:Math><semantics>fs<4fc<annotation encoding="application / x-tex">f_s < 4f_c< / annotation>< / semantics>< / com:Math> since it requires a smaller number of conversion stages as shown in FIGs. 5a and 5b. In this case, the Intermediate Frequency, <com:Math><semantics>fIF<annotation encoding="application / x-tex">f_{IF}< / annotation>< / semantics>< / com:Math>, is selected equal to <com:Math><semantics>fs / 4<annotation encoding="application / x-tex">f_s / 4< / annotation>< / semantics>< / com:Math>. Selection of Carrier Frequency <com:Math><semantics>fc<annotation encoding="application / x-tex">f_c< / annotation>< / semantics>< / com:Math>: In order to select a frequency range s.t. <com:Math><semantics>fc≤fs / 4<annotation encoding="application / x-tex">f_c \le f_s / 4< / annotation>< / semantics>< / com:Math> with relatively low interference and low path loss, while allowing for a multipath-rich environment that is suitable for MIMO communications and while avoiding <com:Math><semantics>ℬHU<annotation encoding="application / x-tex">\mathcal{B}_{HU}< / annotation>< / semantics>< / com:Math>, the disclosure proposes to select <com:Math><semantics>fc∈[2GHz,6GHz]<annotation encoding="application / x-tex">f_c \in [2GHz, 6GHz]< / annotation>< / semantics>< / com:Math>. It is possible to reduce <com:Math><semantics>pk∀k<annotation encoding="application / x-tex">p_k \forall k< / annotation>< / semantics>< / com:Math> by selecting <com:Math><semantics>fc<annotation encoding="application / x-tex">f_c< / annotation>< / semantics>< / com:Math> for the <com:Math><semantics>kth<annotation encoding="application / x-tex">k^{th}< / annotation>< / semantics>< / com:Math> Tx to be distinct from all the other <com:Math><semantics>K−1<annotation encoding="application / x-tex">K-1< / annotation>< / semantics>< / com:Math> carrier frequencies <com:Math><semantics>∀k<annotation encoding="application / x-tex">\forall k< / annotation>< / semantics>< / com:Math>. An optimal selection of the <com:Math><semantics>K<annotation encoding="application / x-tex">K< / annotation>< / semantics>< / com:Math> carrier frequencies is for each frequency to select one unique frequency either from the optimal set <com:Math><semantics>Sfc≜<annotation encoding="application / x-tex">S_{fc} \triangleq< / annotation>< / semantics>< / com:Math> <com:Math><semantics>{fs4κ,3fs4κ,…,(2K−1)fs4κ}<annotation encoding="application / x-tex">\left\{\frac{f_s}{4\kappa}, \frac{3f_s}{4\kappa}, \dots, \frac{(2K-1)f_s}{4\kappa}\right\}< / annotation>< / semantics>< / com:Math> or from any other set such as <com:Math><semantics>Sfc′≜{(K+1)fs8κ,(K+3)fs8κ,…,(3K−1)fs8κ}<annotation encoding="application / x-tex">S'_{f_c} \triangleq \left\{\frac{(K+1)f_s}{8\kappa}, \frac{(K+3)f_s}{8\kappa}, \dots, \frac{(3K-1)f_s}{8\kappa}\right\}< / annotation>< / semantics>< / com:Math>. When <com:Math><semantics>K>1<annotation encoding="application / x-tex">K > 1< / annotation>< / semantics>< / com:Math> and the K carrier frequencies are selected uniquely from <com:Math><semantics>Sfc<annotation encoding="application / x-tex">S_{f_c}< / annotation>< / semantics>< / com:Math> or from <com:Math><semantics>Sfc<annotation encoding="application / x-tex">S_{f_c}< / annotation>< / semantics>< / com:Math>, Post- channel filtering 615 containing an MTF excision filter is recommended at Rx 108, 208, 308 in order to reduce NBI, which originates from the other K-1 interfering carrier frequencies. 5.3.3 MTF System Architecture: The architecture shown in FIGs. 1-4 together with Tx 104, 204, 304 and Rx 108, 208, 308 as described in FIGs. 5a, 5b, 6a, 6b enable an MTF system to communicate with other co-located MTF systems, when using overlapping licensed or unlicensed bands. Such MTF systems include cellular-type, Wi-Fi-type and wireless sensor / Internet-of-Things-type systems, all capable of communicating with one another. FIGs. 1-4 together with FIGs. 5a, 5b, 6a, 6b, have a digital component 510, which includes MTF Modulator 102, 202, 302, a portion of Tx 104, 204, 304, a portion of Rx 108, 208, 308 and MTF detector 110, 210, 310. The digital portion of Tx 104, 204, 304 includes the Pre-Channel filter 500 and the digital portion of Digital-to-Analog (D / A) converter 503. The digital portion of Rx 108, 208, 308 includes the Post-Channel filter 615 and the digital portion of Analog-to-Digital (A / D) converter 613. The digital component is mostly software-defined (S / W) and is allowed to change personality according to the MTF system it is communicating with, by adjusting <com:Math><semantics>{hMTF}k<annotation encoding="application / x-tex">\{h_{MTF}\}_k< / annotation>< / semantics>< / com:Math>, Pre-channel filter 500, Post-channel filter 615 and MMSE-SIC detector 110, 210, 310. Also, the architecture shown in FIGs. 1-4 together with a portion of Tx 104, 204, 304 and a portion of Rx 108, 208, 308 have analog components 511, 513, 616, 618, which are hardware-defined (H / W). As described in FIGs. 5a, 5b, 6a, 6b, the analog portions of Tx 104, 204, 304 and Rx 108, 208, 308 include converters (e.g. D / A 502, Up / Down 504, 611 and A / D 613), analog filters 607 (e.g. Band-Pass Filter (BPF) and Low-Pass Filter (LPF)) and amplifiers (e.g. Power Amplifier (PA) 506 and Low Noise Amplifier (LNA) 609). Given that most existing systems contain a S / W component and a H / W component, it is possible to upgrade such systems to an MTF system through a S / W download, without requiring a H / W modification as long as it is possible to overcome its limitation. For example, when the sampling frequency, <com:Math><semantics>fA / D<annotation encoding="application / x-tex">f_{A / D}< / annotation>< / semantics>< / com:Math>, of the available A / D converter is smaller than the required <com:Math><semantics>fs<annotation encoding="application / x-tex">f_s< / annotation>< / semantics>< / com:Math> by a multiple <com:Math><semantics>𝔇><annotation encoding="application / x-tex">\mathfrak{D} >< / annotation>< / semantics>< / com:Math> 1 s.t. <com:Math><semantics>fA / D=fs / 𝔇<annotation encoding="application / x-tex">f_{A / D} = f_s / \mathfrak{D}< / annotation>< / semantics>< / com:Math>, it is possible to reduce <com:Math><semantics>fs<annotation encoding="application / x-tex">f_s< / annotation>< / semantics>< / com:Math> by <com:Math><semantics>𝔇<annotation encoding="application / x-tex">\mathfrak{D}< / annotation>< / semantics>< / com:Math> to accommodate <com:Math><semantics>fA / D<annotation encoding="application / x-tex">f_{A / D}< / annotation>< / semantics>< / com:Math> while maintaining the same desired channel capacity, <com:Math><semantics>𝒞𝔡<annotation encoding="application / x-tex">\mathcal{C}_{\mathfrak{d}}< / annotation>< / semantics>< / com:Math>, using several non mutually exclusive techniques: MTF Technique 1: Decimate <com:Math><semantics>hMTF<annotation encoding="application / x-tex">h_{MTF}< / annotation>< / semantics>< / com:Math> by <com:Math><semantics>𝔇<annotation encoding="application / x-tex">\mathfrak{D}< / annotation>< / semantics>< / com:Math>, while increasing <com:Math><semantics>Ts<annotation encoding="application / x-tex">T_s< / annotation>< / semantics>< / com:Math> by <com:Math><semantics>𝔇<annotation encoding="application / x-tex">\mathfrak{D}< / annotation>< / semantics>< / com:Math>. MTF Technique 2: Relax Constraint 3 by loading <com:Math><semantics>𝔇>1<annotation encoding="application / x-tex">\mathfrak{D} > 1< / annotation>< / semantics>< / com:Math> bits of information / DOF. MTF Technique 3: Select <com:Math><semantics>Kb=𝔇<annotation encoding="application / x-tex">K_b = \mathfrak{D}< / annotation>< / semantics>< / com:Math> while forcing each column of <com:Math><semantics>Kb<annotation encoding="application / x-tex">K_b< / annotation>< / semantics>< / com:Math> to have a distinct <com:Math><semantics>fc<annotation encoding="application / x-tex">f_c< / annotation>< / semantics>< / com:Math>. It is possible to combine several of the MTF techniques shown above in order to overcome the limitation of having <com:Math><semantics>fA / D=fs / 𝔇<annotation encoding="application / x-tex">f_{A / D} = f_s / \mathfrak{D}< / annotation>< / semantics>< / com:Math>. For example, by increasing the number of information bits / DOF from 1 bit to 2 bits, while increasing <com:Math><semantics>K𝔡<annotation encoding="application / x-tex">K_{\mathfrak{d}}< / annotation>< / semantics>< / com:Math> from 1 to 4, we have <com:Math><semantics>𝔇=8<annotation encoding="application / x-tex">\mathfrak{D} = 8< / annotation>< / semantics>< / com:Math>. 5.4 MTF MA NETWORKS This section designs MTF MA networks across a centralized topology similar to existing MA networks, such as LTE and Wi-Fi networks. As typical of any centralized topology, the MTF MA network consists of two types of transmissions: (a) downlink (DL) transmissions, from Base Station (BS) or Access Point (AP) to device; and (b) uplink (UL) transmissions, from device to BS / AP. The designs of the MTF MA networks are based on the following assumptions: 5.4.1 Assumptions: a) Several co-located centralized MTF MA networks use overlapping licensed or unlicensed bands. Based on the system architecture in Section 5.3.3, such networks are capable to collaborate, which offers many benefits. For example Time Division Duplex (TDD) can be implemented, which forces temporal separation between DL and UL transmissions. b) The MTF BS / AP contains an antenna array to be used in beamforming in the DL portion of most of the MTF MA networks in order to reduce WBI from interfering BSs / APs at devices. c) The range, <com:Math><semantics>ℜk<annotation encoding="application / x-tex">\Re_k< / annotation>< / semantics>< / com:Math>, between the <com:Math><semantics>kth<annotation encoding="application / x-tex">k^{th}< / annotation>< / semantics>< / com:Math> MTF Tx 104, 304 and an MTF Rx 108, 308 is a function of the link budget, <com:Math><semantics>𝔔k<annotation encoding="application / x-tex">\mathfrak{Q}_k< / annotation>< / semantics>< / com:Math>, between them, where [Image disponible dans le document PDF, Image available in the PDF document] (21) <com:Math><semantics>𝔭dBm<annotation encoding="application / x-tex">\mathfrak{p}_{dBm}< / annotation>< / semantics>< / com:Math> is the average transmitted power; <com:Math><semantics>NFdB<annotation encoding="application / x-tex">NF_{dB}< / annotation>< / semantics>< / com:Math> is the Noise Figure of Rx; <com:Math><semantics>𝔭k𝒫<annotation encoding="application / x-tex">\mathfrak{p}_k^{\mathcal{P}}< / annotation>< / semantics>< / com:Math> corresponds to <com:Math><semantics>fs𝒫≜fs(1−ζex)<annotation encoding="application / x-tex">f_s^{\mathcal{P}} \triangleq f_s(1-\zeta_{ex})< / annotation>< / semantics>< / com:Math>, which is the Noise-equivalent BW in Hz at the output of the post-channel filter; <com:Math><semantics>ζex<annotation encoding="application / x-tex">\zeta_{ex}< / annotation>< / semantics>< / com:Math> is the excision factor, defined as <com:Math><semantics>ζex≜1fc∫f∈ℬexfdf<annotation encoding="application / x-tex">\zeta_{ex} \triangleq \frac{1}{f_c} \int_{f \in \mathcal{B}_{ex}} f \, df< / annotation>< / semantics>< / com:Math>; <com:Math><semantics>GdB<annotation encoding="application / x-tex">G_{dB}< / annotation>< / semantics>< / com:Math> is the antenna gain between Tx and Rx. d) A unique <com:Math><semantics>g→l,k<annotation encoding="application / x-tex">\vec{g}_{l,k}< / annotation>< / semantics>< / com:Math> 1203, 1206 as defined in (17) and a unique <com:Math><semantics>℘→k,0<annotation encoding="application / x-tex">\vec{\wp}_{k,0}< / annotation>< / semantics>< / com:Math> 1209 as defined in (16) are selected for the <com:Math><semantics>kth<annotation encoding="application / x-tex">k^{th}< / annotation>< / semantics>< / com:Math> MTF device <com:Math><semantics>∀k<annotation encoding="application / x-tex">\forall k< / annotation>< / semantics>< / com:Math>. e) <com:Math><semantics>℘→k,ℵk𝒫<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,\aleph_k}^{\mathcal{P}}< / annotation>< / semantics>< / com:Math> 1208, is designed s.t. its PSD, <com:Math><semantics>𝒮℘k,ℵk𝒫(t)(f)<annotation encoding="application / x-tex">\mathcal{S}_{\wp_{k,\aleph_k}^{\mathcal{P}}(t)}(f)< / annotation>< / semantics>< / com:Math> 808, has nulls at <com:Math><semantics>ℬNBI∪ℬex<annotation encoding="application / x-tex">\mathcal{B}_{NBI} \cup \mathcal{B}_{ex}< / annotation>< / semantics>< / com:Math> where <com:Math><semantics>℘k,ℵk𝒫(t)<annotation encoding="application / x-tex">\wp_{k,\aleph_k}^{\mathcal{P}}(t)< / annotation>< / semantics>< / com:Math> is the continuous-time version of <com:Math><semantics>℘→k,ℵν𝒫<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,\aleph_{\nu}}^{\mathcal{P}}< / annotation>< / semantics>< / com:Math> 1208. f) <com:Math><semantics>℘→k,0𝒫<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,0}^{\mathcal{P}}< / annotation>< / semantics>< / com:Math> 1209 is designed s.t. its PSD, <com:Math><semantics>𝒮℘k,0𝒫(t)(f)<annotation encoding="application / x-tex">\mathcal{S}_{\wp_{k,0}^{\mathcal{P}}(t)}(f)< / annotation>< / semantics>< / com:Math> 809, is pre-distorted according to (19), where <com:Math><semantics>𝒟k,0𝒫(t)<annotation encoding="application / x-tex">\mathscr{D}_{k,0}^{\mathscr{P}}(t)< / annotation>< / semantics>< / com:Math> is the continuous-time version of <com:Math><semantics>𝒟→k,0𝒫<annotation encoding="application / x-tex">\overrightarrow{\mathscr{D}}_{k,0}^{\mathscr{P}}< / annotation>< / semantics>< / com:Math> 1209. g) <com:Math><semantics>d|UL=N<annotation encoding="application / x-tex">d|_{UL} = N< / annotation>< / semantics>< / com:Math> in the UL portion of an MTF MA network in order to decrease <com:Math><semantics>max⁡k{pk}<annotation encoding="application / x-tex">\max_{k} \{p_k\}< / annotation>< / semantics>< / com:Math> in (21), while <com:Math><semantics>d|DL=1<annotation encoding="application / x-tex">d|_{DL} = 1< / annotation>< / semantics>< / com:Math> in the DL portion in order to maintain a high DL capacity, <com:Math><semantics>𝒞MTF<annotation encoding="application / x-tex">\mathcal{C}_{MTF}< / annotation>< / semantics>< / com:Math>, where <com:Math><semantics>d|UL<annotation encoding="application / x-tex">d|_{UL}< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>d|DL<annotation encoding="application / x-tex">d|_{DL}< / annotation>< / semantics>< / com:Math> are the delays in (2) corresponding to UL and DL respectively. An additional reason for selecting <com:Math><semantics>d|UL=N<annotation encoding="application / x-tex">d|_{UL} = N< / annotation>< / semantics>< / com:Math> is to reduce the Peak-to-Average Power Ratio (PAPR) corresponding to transmissions from an MTF device, which can be reduced even further by selecting <com:Math><semantics>℘k,ℵk<annotation encoding="application / x-tex">\wp_{k,\aleph_k}< / annotation>< / semantics>< / com:Math> 1208 in (9) as <com:Math><semantics>℘→REC,1<annotation encoding="application / x-tex">\overrightarrow{\wp}_{REC,1}< / annotation>< / semantics>< / com:Math> 1214. A further reason for selecting <com:Math><semantics>d|UL=N<annotation encoding="application / x-tex">d|_{UL} = N< / annotation>< / semantics>< / com:Math> is to have a memoryless MTF MA network with <com:Math><semantics>𝒞MTF|L<∞=𝒞MTF|L→∞<annotation encoding="application / x-tex">\mathcal{C}_{MTF}|_{L<\infty} = \mathcal{C}_{MTF}|_{L\to\infty}< / annotation>< / semantics>< / com:Math>. On the other hand, selecting <com:Math><semantics>𝒹|DL=1<annotation encoding="application / x-tex">\mathcal{d}|_{DL} = 1< / annotation>< / semantics>< / com:Math> in the DL portion implies that the MTF MA network has memory and <com:Math><semantics>𝒞MTF|L<∞=L / (L+Nd−<annotation encoding="application / x-tex">\mathcal{C}_{MTF}|_{L<\infty} = L / (L + \frac{N}{d} -< / annotation>< / semantics>< / com:Math> 1) <com:Math><semantics>𝒞MTF<annotation encoding="application / x-tex">\mathcal{C}_{MTF}< / annotation>< / semantics>< / com:Math>. For example, when <com:Math><semantics>10N<L<∞<annotation encoding="application / x-tex">10N < L < \infty< / annotation>< / semantics>< / com:Math>, <com:Math><semantics>𝒞MTF|L<∞>0.9𝒞MTF|L→∞<annotation encoding="application / x-tex">\mathcal{C}_{MTF}|_{L < \infty} > 0.9 \mathcal{C}_{MTF}|_{L \to \infty}< / annotation>< / semantics>< / com:Math>. In the UL portion, all Q symbols in <com:Math><semantics>α→<annotation encoding="application / x-tex">\vec{\alpha}< / annotation>< / semantics>< / com:Math> 101, 201, 301, corresponding to all the K active Txs 104, 204, 304, are required to be detected, while in the DL portion, only the desired symbols in <com:Math><semantics>α→<annotation encoding="application / x-tex">\vec{\alpha}< / annotation>< / semantics>< / com:Math> 101, 201, 301, corresponding to the desired Tx 104, 204, 304, are required to be detected, with the remaining symbols, corresponding to the <com:Math><semantics>Ki=K−1<annotation encoding="application / x-tex">K_i = K - 1< / annotation>< / semantics>< / com:Math> interfering columns in <com:Math><semantics>hMTF<annotation encoding="application / x-tex">h_{MTF}< / annotation>< / semantics>< / com:Math> 1104, ignored. For this reason, a preferred embodiment is to constrain <com:Math><semantics>max⁡k{pk}<annotation encoding="application / x-tex">\max_{k} \{p_k\}< / annotation>< / semantics>< / com:Math> in (21) to correspond to a full implementation of Constraint <com:Math><semantics>3∀k<annotation encoding="application / x-tex">3 \forall k< / annotation>< / semantics>< / com:Math> for the UL portion, while in the DL portion, a preferred embodiment is to constrain <com:Math><semantics>max⁡k{pk}<annotation encoding="application / x-tex">\max_{k} \{p_k\}< / annotation>< / semantics>< / com:Math> in (21) to correspond to a partial implementation of Constraint 3 corresponding only to the desired received symbols in <com:Math><semantics>α→<annotation encoding="application / x-tex">\vec{\alpha}< / annotation>< / semantics>< / com:Math> 101, 201, 301. 5.4.2 Designs of MTF MA networks: Based on the above assumptions, we design 3 MTF MA networks, namely <com:Math><semantics>MTF1<annotation encoding="application / x-tex">MTF_1< / annotation>< / semantics>< / com:Math>, <com:Math><semantics>MTF2<annotation encoding="application / x-tex">MTF_2< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>MTF3<annotation encoding="application / x-tex">MTF_3< / annotation>< / semantics>< / com:Math>, all constrained by a mask with a BW, <com:Math><semantics>𝒲m=20MHz<annotation encoding="application / x-tex">\mathcal{W}_m = 20MHz< / annotation>< / semantics>< / com:Math>. This implies that <com:Math><semantics>(NTs)min≜<annotation encoding="application / x-tex">(NT_s)_{min} \triangleq< / annotation>< / semantics>< / com:Math> <com:Math><semantics>r / 20MHz<annotation encoding="application / x-tex">r / 20MHz< / annotation>< / semantics>< / com:Math> 1301. For example, when <com:Math><semantics>ℙMask(f)<annotation encoding="application / x-tex">\mathbb{P}_{Mask}(f)< / annotation>< / semantics>< / com:Math> 701 is selected as <com:Math><semantics>ℙWiFi(f)<annotation encoding="application / x-tex">\mathbb{P}_{WiFi}(f)< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>℘→k,ℵk𝒫<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,\aleph_k}^{\mathcal{P}}< / annotation>< / semantics>< / com:Math> 1208 as <com:Math><semantics>𝒟→REC,1𝒫<annotation encoding="application / x-tex">\overrightarrow{\mathcal{D}}_{REC,1}^{\mathcal{P}}< / annotation>< / semantics>< / com:Math> 1214 with a PSD <com:Math><semantics>𝒮𝒟REC,1𝒫(t)(f)<annotation encoding="application / x-tex">\mathcal{S}_{\mathcal{D}_{REC,1}^{\mathcal{P}}(t)}(f)< / annotation>< / semantics>< / com:Math> 908, <com:Math><semantics>r=64 / 3=21.3<annotation encoding="application / x-tex">r = 64 / 3 = 21.3< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>(NTs)min≜r / 20MHz=1.3<annotation encoding="application / x-tex">(NT_s)_{min} \triangleq r / 20MHz = 1.3< / annotation>< / semantics>< / com:Math> 1.066<com:Math><semantics>μ<annotation encoding="application / x-tex">\mu< / annotation>< / semantics>< / com:Math>s 1301. On the other hand, when <com:Math><semantics>ℙMask(f)<annotation encoding="application / x-tex">\mathbb{P}_{Mask}(f)< / annotation>< / semantics>< / com:Math> 107 is selected as <com:Math><semantics>ℙLTE(f)<annotation encoding="application / x-tex">\mathbb{P}_{LTE}(f)< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>℘→k,ℵk𝒫<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,\aleph_k}^{\mathcal{P}}< / annotation>< / semantics>< / com:Math> 1208 as <com:Math><semantics>℘→REC,1𝒫<annotation encoding="application / x-tex">\overrightarrow{\wp}_{REC,1}^{\mathcal{P}}< / annotation>< / semantics>< / com:Math> 1214 with a PSD <com:Math><semantics>S℘REC,1𝒫(t)(f)<annotation encoding="application / x-tex">S_{\wp_{REC,1}^{\mathcal{P}}(t)}(f)< / annotation>< / semantics>< / com:Math> 908, <com:Math><semantics>𝒱=40.8<annotation encoding="application / x-tex">\mathscr{V} = 40.8< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>(NTs)min≜𝒱20MHz=2.04μs<annotation encoding="application / x-tex">(NT_s)_{min} \triangleq \frac{\mathscr{V}}{20MHz} = 2.04\mu s< / annotation>< / semantics>< / com:Math> 1301. Moreover, since the DL portion of each network is assumed to have relatively low interference, it is characterized to be with memory, with <com:Math><semantics>d=K=1<annotation encoding="application / x-tex">d = K = 1< / annotation>< / semantics>< / com:Math>, and <com:Math><semantics>Ts,max|DI≜1 / Cb|DL≥1 / fs|DL<annotation encoding="application / x-tex">T_{s,max}|_{DI} \triangleq 1 / C_b|_{DL} \geq 1 / f_s|_{DL}< / annotation>< / semantics>< / com:Math> 1301. On the other hand, since the UL portion of each network is assumed to have relatively high interference, it is characterized as memoryless, with <com:Math><semantics>d=N,K≫1<annotation encoding="application / x-tex">d = N, K \gg 1< / annotation>< / semantics>< / com:Math>, and <com:Math><semantics>Ts,max|UL≜κN|𝒞b|UL≥1<annotation encoding="application / x-tex">T_{s,max}|_{UL} \triangleq \frac{\kappa}{N |\mathcal{C}_b|_{UL}} \geq 1< / annotation>< / semantics>< / com:Math> <com:Math><semantics>1 / fS|UL<annotation encoding="application / x-tex">1 / f_S|_{UL}< / annotation>< / semantics>< / com:Math> 1301. Therefore, <com:Math><semantics>Nmin|DL≜(NTS)minTSmax|DL=r𝒞b|DL20MHz<annotation encoding="application / x-tex">N_{min}|_{DL} \triangleq \frac{(NT_S)_{min}}{T_{Smax}|_{DL}} = \frac{r \ \mathcal{C}_b|_{DL}}{20MHz}< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>Nmin|UL≜(NTS)minTSmax|UL=r𝒞b|ULN20MHzK<annotation encoding="application / x-tex">N_{min}|_{UL} \triangleq \frac{(NT_S)_{min}}{T_{Smax}|_{UL}} = \frac{r \ \mathcal{C}_b|_{UL} N}{20MHz \ K}< / annotation>< / semantics>< / com:Math> 1301. Design Parameters for <com:Math><semantics>MTF1<annotation encoding="application / x-tex">MTF_1< / annotation>< / semantics>< / com:Math>, <com:Math><semantics>MTF2<annotation encoding="application / x-tex">MTF_2< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>MTF3<annotation encoding="application / x-tex">MTF_3< / annotation>< / semantics>< / com:Math>: 1. <com:Math><semantics>MTF1<annotation encoding="application / x-tex">MTF_1< / annotation>< / semantics>< / com:Math> is selected to have a desired DL channel capacity of <com:Math><semantics>𝒞𝔡1|DL=0.2Gbps<annotation encoding="application / x-tex">\mathcal{C}_{\mathfrak{d}_1}|_{DL} = 0.2 Gbps< / annotation>< / semantics>< / com:Math> 1301, and a desired UL channel capacity of <com:Math><semantics>𝒞𝔡1|UL=0.2Gbps<annotation encoding="application / x-tex">\mathcal{C}_{\mathfrak{d}_1}|_{UL} = 0.2 Gbps< / annotation>< / semantics>< / com:Math> 1301, both across the unlicensed (Title 47) CFR 15.247) mid-band frequency of <com:Math><semantics>fc1=2.45GHz<annotation encoding="application / x-tex">f_{c_1} = 2.45 GHz< / annotation>< / semantics>< / com:Math>. We also select for both DL and UL, <com:Math><semantics>|fs1|DL=|fs1|UL=0.2GHz<annotation encoding="application / x-tex">\left|f_{s_1}\right|_{DL} = \left|f_{s_1}\right|_{UL} = 0.2GHz< / annotation>< / semantics>< / com:Math> and IF sampling with <com:Math><semantics>ℬTL1≜|fc1−fs1 / 4,fc1+fs14|<annotation encoding="application / x-tex">\mathcal{B}_{TL_1} \triangleq \left|f_{c_1} - f_{s_1} / 4, f_{c_1} + \frac{f_{s_1}}{4}\right|< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>ℬex1≜|fc1−fs2 / 4,fc1+fs24|<annotation encoding="application / x-tex">\mathcal{B}_{ex_1} \triangleq \left|f_{c_1} - f_{s_2} / 4, f_{c_1} + \frac{f_{s_2}}{4}\right|< / annotation>< / semantics>< / com:Math> <com:Math><semantics>{{ℬNBI1∪ℬRB∪ℬHU}∩ℬTL1}∪ℬ¯TL1.<annotation encoding="application / x-tex">\left\{ \left\{ \mathcal{B}_{NBI_1} \cup \mathcal{B}_{RB} \cup \mathcal{B}_{HU} \right\} \cap \mathcal{B}_{TL_1} \right\} \cup \overline{\mathcal{B}}_{TL_1}.< / annotation>< / semantics>< / com:Math> 2. <com:Math><semantics>MTF2<annotation encoding="application / x-tex">MTF_2< / annotation>< / semantics>< / com:Math> is selected to have <com:Math><semantics>𝒞𝔡2|DL=2Gbps𝟏𝟑𝟎𝟏<annotation encoding="application / x-tex">\mathcal{C}_{\mathfrak{d}_2}|_{DL} = 2Gbps \ \mathbf{1301}< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>𝒞𝔡2|UL=2Gbps𝟏𝟑𝟎𝟏<annotation encoding="application / x-tex">\mathcal{C}_{\mathfrak{d}_2}|_{UL} = 2Gbps \ \mathbf{1301}< / annotation>< / semantics>< / com:Math>, both across the unlicensed mid-band frequency of <com:Math><semantics>fc2=5.8GHz<annotation encoding="application / x-tex">f_{c_2} = 5.8 GHz< / annotation>< / semantics>< / com:Math>. We also select <com:Math><semantics>fs2|DL=fs2|UL=2GHz<annotation encoding="application / x-tex">f_{s_2}|_{DL} = f_{s_2}|_{UL} = 2 GHz< / annotation>< / semantics>< / com:Math> and IF sampling with <com:Math><semantics>ℬTL2≜[fc2−fs2 / 4,fc2+fs2 / 4]<annotation encoding="application / x-tex">\mathcal{B}_{TL_2} \triangleq \left[ f_{c_2} - f_{s_2} / 4, f_{c_2} + f_{s_2} / 4 \right]< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>ℬex2≜{{ℬNBI2∪ℬRB∪ℬHU}∩}<annotation encoding="application / x-tex">\mathcal{B}_{ex_2} \triangleq \left\{ \left\{ \mathcal{B}_{NBI_2} \cup \mathcal{B}_{RB} \cup \mathcal{B}_{HU} \right\} \cap \right\}< / annotation>< / semantics>< / com:Math> <com:Math><semantics>ℬTL2<annotation encoding="application / x-tex">\mathcal{B}_{TL_2}< / annotation>< / semantics>< / com:Math> <com:Math><semantics>∪ℬ¯TL2<annotation encoding="application / x-tex">\cup \overline{\mathcal{B}}_{TL_2}< / annotation>< / semantics>< / com:Math>. 3. <com:Math><semantics>MTF3<annotation encoding="application / x-tex">MTF_3< / annotation>< / semantics>< / com:Math> uses a licensed band at <com:Math><semantics>fc3=3.5GHz<annotation encoding="application / x-tex">f_{c_3} = 3.5 GHz< / annotation>< / semantics>< / com:Math> with <com:Math><semantics>𝒞𝔡3|DL=14Gbps<annotation encoding="application / x-tex">\mathcal{C}_{\mathfrak{d}_3}\big|_{DL} = 14 Gbps< / annotation>< / semantics>< / com:Math> 1301 and <com:Math><semantics>𝒞𝔡1|UL=6Gbps<annotation encoding="application / x-tex">\mathcal{C}_{\mathfrak{d}_1}\big|_{UL} = 6 Gbps< / annotation>< / semantics>< / com:Math> 1301. We also select <com:Math><semantics>fs3|DL=fs3|UL=14GHz<annotation encoding="application / x-tex">f_{s_3}|_{DL} = f_{s_3}|_{UL} = 14GHz< / annotation>< / semantics>< / com:Math> and RF sampling with <com:Math><semantics>ℬTL3≜[fc3−fc3]UL=fs3|UL=fs3|UL=fs3|UL=fs3|UL=fs3|UL=fs3|UL=<annotation encoding="application / x-tex">\mathcal{B}_{TL_3} \triangleq [f_{c_3} - f_{c_3}]_{UL} = f_{s_3}|_{UL} = f_{s_3}|_{UL} = f_{s_3}|_{UL} = f_{s_3}|_{UL} = f_{s_3}|_{UL} = f_{s_3}|_{UL} =< / annotation>< / semantics>< / com:Math> <com:Math><semantics>fs3 / 4<annotation encoding="application / x-tex">f_{s_3} / 4< / annotation>< / semantics>< / com:Math>, <com:Math><semantics>fc3+fs3 / 4<annotation encoding="application / x-tex">f_{c_3}+f_{s_3} / 4< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>ℬex3≜{{ℬNBI3∪ℬRB∪ℬHU}∩ℬTL3}∪ℬ¯TL3<annotation encoding="application / x-tex">\mathcal{B}_{ex_3} \triangleq \left\{ \left\{ \mathcal{B}_{NBI_3} \cup \mathcal{B}_{RB} \cup \mathcal{B}_{HU} \right\} \cap \mathcal{B}_{TL_3} \right\} \cup \overline{\mathcal{B}}_{TL_3}< / annotation>< / semantics>< / com:Math>. FIG. 9 displays the PSD, <com:Math><semantics>𝒮℘REC,1𝒫(t)(f)<annotation encoding="application / x-tex">\mathcal{S}_{\wp_{REC,1}^{\mathcal{P}}(t)}(f)< / annotation>< / semantics>< / com:Math> 908, of <com:Math><semantics>℘→REC,1𝒫<annotation encoding="application / x-tex">\overrightarrow{\wp}_{REC,1}^{\mathcal{P}}< / annotation>< / semantics>< / com:Math> and the PSD, <com:Math><semantics>𝒮℘k,0𝒫(t)(f)<annotation encoding="application / x-tex">\mathcal{S}_{\wp_{k,0}^{\mathcal{P}}(t)}(f)< / annotation>< / semantics>< / com:Math> 809, of <com:Math><semantics>℘→k,0𝒫<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,0}^{\mathcal{P}}< / annotation>< / semantics>< / com:Math> 1209 for <com:Math><semantics>MTF3<annotation encoding="application / x-tex">MTF_3< / annotation>< / semantics>< / com:Math>, with nulls at <com:Math><semantics>ℬex3<annotation encoding="application / x-tex">\mathcal{B}_{ex_3}< / annotation>< / semantics>< / com:Math>, when <com:Math><semantics>℘→k,ℵk𝒫<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,\aleph_k}^{\mathcal{P}}< / annotation>< / semantics>< / com:Math> 1208 is selected as <com:Math><semantics>℘→REC,1𝒫<annotation encoding="application / x-tex">\overrightarrow{\wp}_{REC,1}^{\mathcal{P}}< / annotation>< / semantics>< / com:Math> 1214 while <com:Math><semantics>℘→k,0𝒫<annotation encoding="application / x-tex">\overrightarrow{\wp}_{k,0}^{\mathcal{P}}< / annotation>< / semantics>< / com:Math> 1209 is selected as defined in (16) pre-distorted according to (19). One advantage of <com:Math><semantics>MTF3<annotation encoding="application / x-tex">MTF_3< / annotation>< / semantics>< / com:Math> over <com:Math><semantics>MTF1<annotation encoding="application / x-tex">MTF_1< / annotation>< / semantics>< / com:Math> and over <com:Math><semantics>MTF2<annotation encoding="application / x-tex">MTF_2< / annotation>< / semantics>< / com:Math> is that it is possible to increase <com:Math><semantics>𝒞𝔡3|DL<annotation encoding="application / x-tex">\mathcal{C}_{\mathfrak{d}_3}\big|_{DL}< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>𝒞𝔡3|HL<annotation encoding="application / x-tex">\mathcal{C}_{\mathfrak{d}_3}\big|_{HL}< / annotation>< / semantics>< / com:Math> by a multiple <com:Math><semantics>𝔇<annotation encoding="application / x-tex">\mathfrak{D}< / annotation>< / semantics>< / com:Math> by decreasing <com:Math><semantics>Ts≜1 / fs3<annotation encoding="application / x-tex">T_s \triangleq 1 / f_{s_3}< / annotation>< / semantics>< / com:Math> by <com:Math><semantics>𝔇<annotation encoding="application / x-tex">\mathfrak{D}< / annotation>< / semantics>< / com:Math> regardless of the value of <com:Math><semantics>𝒲m<annotation encoding="application / x-tex">\mathcal{W}_m< / annotation>< / semantics>< / com:Math> at the cost of an increase in <com:Math><semantics>pk<annotation encoding="application / x-tex">p_k< / annotation>< / semantics>< / com:Math>. Practical Consideration I: In general, it is possible to increase <com:Math><semantics>Cb|UL<annotation encoding="application / x-tex">C_b|_{UL}< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>Cb|DL<annotation encoding="application / x-tex">C_b|_{DL}< / annotation>< / semantics>< / com:Math> by a multiple <com:Math><semantics>𝔇<annotation encoding="application / x-tex">\mathfrak{D}< / annotation>< / semantics>< / com:Math> while maintaining the same <com:Math><semantics>fs<annotation encoding="application / x-tex">f_s< / annotation>< / semantics>< / com:Math> by selecting any combination of the 3 MTF Techniques 1-3 described in Section 5.3.3. For example, by increasing the number of information bits / DOF from 1 bit to 2 bits / DOF, while increasing <com:Math><semantics>Kb<annotation encoding="application / x-tex">K_b< / annotation>< / semantics>< / com:Math> from 1 to 4, we can have <com:Math><semantics>𝔇=8<annotation encoding="application / x-tex">\mathfrak{D} = 8< / annotation>< / semantics>< / com:Math>, and consequently, <com:Math><semantics>𝒞𝔡1|DL=𝒞𝔡1|DL=1.6Gbps,𝒞𝔡2|DL=𝒞𝔡2|DL=16Gbps and 𝒞𝔡3|DL=𝒞𝔡3|DL=112Gbps.<annotation encoding="application / x-tex">\left. \mathcal{C}_{\mathfrak{d}_1} \right|_{DL} = \left. \mathcal{C}_{\mathfrak{d}_1} \right|_{DL} = 1.6Gbps, \left. \mathcal{C}_{\mathfrak{d}_2} \right|_{DL} = \left. \mathcal{C}_{\mathfrak{d}_2} \right|_{DL} = 16Gbps \text{ and } \left. \mathcal{C}_{\mathfrak{d}_3} \right|_{DL} = \left. \mathcal{C}_{\mathfrak{d}_3} \right|_{DL} = 112Gbps.< / annotation>< / semantics>< / com:Math> Practical Consideration II: It is possible to increase K by a multiple <com:Math><semantics>𝔇<annotation encoding="application / x-tex">\mathfrak{D}< / annotation>< / semantics>< / com:Math> while keeping <com:Math><semantics>𝒞𝔡|UL<annotation encoding="application / x-tex">\mathcal{C}_{\mathfrak{d}}|_{UL}< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>C𝔡|DL<annotation encoding="application / x-tex">C_{\mathfrak{d}}|_{DL}< / annotation>< / semantics>< / com:Math> fixed and while maintaining the same <com:Math><semantics>fs<annotation encoding="application / x-tex">f_s< / annotation>< / semantics>< / com:Math>, by selecting MTF Technique 3 in Section 5.3.3. In this case, the increase in <com:Math><semantics>pk<annotation encoding="application / x-tex">p_k< / annotation>< / semantics>< / com:Math> is reasonable as long as <com:Math><semantics>𝔇≤8<annotation encoding="application / x-tex">\mathfrak{D} \leq 8< / annotation>< / semantics>< / com:Math>. If <com:Math><semantics>𝔇>8<annotation encoding="application / x-tex">\mathfrak{D} > 8< / annotation>< / semantics>< / com:Math>, the increase in <com:Math><semantics>pk<annotation encoding="application / x-tex">p_k< / annotation>< / semantics>< / com:Math> can remain reasonable by increasing the delay, <com:Math><semantics>d<annotation encoding="application / x-tex">d< / annotation>< / semantics>< / com:Math>. For example, it is possible to double <com:Math><semantics>𝔇<annotation encoding="application / x-tex">\mathfrak{D}< / annotation>< / semantics>< / com:Math> from 8 to 16, by doubling d from 1 to 2 samples, which halves <com:Math><semantics>Cb|UL<annotation encoding="application / x-tex">C_b|_{UL}< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>Cb|DL<annotation encoding="application / x-tex">C_b|_{DL}< / annotation>< / semantics>< / com:Math>. The implication of having K increase from 1 to 16 implies that 16 co-located MTF networks can co-exist across the same overlapping licensed and unlicensed bands, after forcing <com:Math><semantics>Cδ|UL<annotation encoding="application / x-tex">C_{\delta}|_{UL}< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>Cδ|DL<annotation encoding="application / x-tex">C_{\delta}|_{DL}< / annotation>< / semantics>< / com:Math> to be halved. Practical Consideration III: It is possible to increase K by <com:Math><semantics>𝔇>8<annotation encoding="application / x-tex">\mathfrak{D} > 8< / annotation>< / semantics>< / com:Math> without forcing <com:Math><semantics>𝒞𝔡|UL<annotation encoding="application / x-tex">\mathcal{C}_{\mathfrak{d}}|_{UL}< / annotation>< / semantics>< / com:Math> and <com:Math><semantics>C𝔡|DL<annotation encoding="application / x-tex">C_{\mathfrak{d}}|_{DL}< / annotation>< / semantics>< / com:Math> to be halved by decreasing the ratio, <com:Math><semantics>K / Nex¯<annotation encoding="application / x-tex">K / N_{\overline{ex}}< / annotation>< / semantics>< / com:Math>, where <com:Math><semantics>Nex¯≜N(1−ζex)<annotation encoding="application / x-tex">N_{\overline{ex}} \triangleq N(1-\zeta_{ex})< / annotation>< / semantics>< / com:Math>, as shown below: Increasing <com:Math><semantics>Nex¯<annotation encoding="application / x-tex">N_{\overline{ex}}< / annotation>< / semantics>< / com:Math> when <com:Math><semantics>K<annotation encoding="application / x-tex">K< / annotation>< / semantics>< / com:Math> is fixed: a) One way to increase <com:Math><semantics>Nex¯<annotation encoding="application / x-tex">N_{\overline{ex}}< / annotation>< / semantics>< / com:Math> is by increasing <com:Math><semantics>Nℐ<annotation encoding="application / x-tex">N_{\mathcal{I}}< / annotation>< / semantics>< / com:Math> at Tx 104, 204, 304, which does not affect the desired DL capacity per MTF device, <com:Math><semantics>Cb|DL<annotation encoding="application / x-tex">C_b|_{DL}< / annotation>< / semantics>< / com:Math> 1301. It reduces the capacity per MTF device for the UL portion, without affecting the overall desired network capacity, <com:Math><semantics>Cb|UL<annotation encoding="application / x-tex">C_b|_{UL}< / annotation>< / semantics>< / com:Math> 1301. b) Another way to increase <com:Math><semantics>Nex<annotation encoding="application / x-tex">N_{ex}< / annotation>< / semantics>< / com:Math> is by increasing <com:Math><semantics>NII<annotation encoding="application / x-tex">N_{II}< / annotation>< / semantics>< / com:Math> indirectly at Rx 108, 208, 308 by taking advantage of the frequency-selective nature of a multipath-rich communication channel 106, 306, 406, which forces N to be replaced by <com:Math><semantics>𝒩≥N<annotation encoding="application / x-tex">\mathcal{N} \geq N< / annotation>< / semantics>< / com:Math>, or equivalently, forces <com:Math><semantics>Nex¯<annotation encoding="application / x-tex">N_{\overline{ex}}< / annotation>< / semantics>< / com:Math> to be replaced by <com:Math><semantics>𝒩ex≜𝒩(1−ζex)<annotation encoding="application / x-tex">\mathcal{N}_{ex} \triangleq \mathcal{N}(1-\zeta_{ex})< / annotation>< / semantics>< / com:Math>. This is often referred to as multipath diversity. In this case, <com:Math><semantics>g→l,k<annotation encoding="application / x-tex">\vec{g}_{l,k}< / annotation>< / semantics>< / com:Math> 1203, 1206 in (17) is replaced by <com:Math><semantics>g→l,k,Ch<annotation encoding="application / x-tex">\vec{g}_{l,k,Ch}< / annotation>< / semantics>< / com:Math> in (18) at Rx 108, 308. C) Another way to increase <com:Math><semantics>Nex<annotation encoding="application / x-tex">N_{ex}< / annotation>< / semantics>< / com:Math> indirectly at Rx 108, 208, 308 is by increasing the number, <com:Math><semantics>Nr<annotation encoding="application / x-tex">N_r< / annotation>< / semantics>< / com:Math>, of receive antennas, which does not affect the channel capacity per MTF device, while it increases the overall network capacity accordingly. This is often referred to as spatial diversity or MU-MIMO as shown in FIG. 4. Assuming that each MTF BS / AP contains an array of <com:Math><semantics>Nr<annotation encoding="application / x-tex">N_r< / annotation>< / semantics>< / com:Math> receiving antennas and that K is the number of simultaneously active MTF devices, therefore, the <com:Math><semantics>ith<annotation encoding="application / x-tex">i^{th}< / annotation>< / semantics>< / com:Math> receiving antenna produces the received signal, <com:Math><semantics>r→i≜y→i+w→i∀i∈<annotation encoding="application / x-tex">\vec{r}_i \triangleq \vec{y}_i + \vec{w}_i \ \forall i \in< / annotation>< / semantics>< / com:Math> <com:Math><semantics>{1,…,Nr}<annotation encoding="application / x-tex">\{1, \dots, N_r\}< / annotation>< / semantics>< / com:Math> where <com:Math><semantics>y→i≜h→i,1,Ch*β→1+…+h→i,K,Ch*β→K;h→i,j,Ch<annotation encoding="application / x-tex">\vec{y}_i \triangleq \vec{h}_{i,1,Ch} * \vec{\beta}_1 + \dots + \vec{h}_{i,K,Ch} * \vec{\beta}_K; \vec{h}_{i,j,Ch}< / annotation>< / semantics>< / com:Math> is the communications channel 106, 306, 406 between the <com:Math><semantics>jth<annotation encoding="application / x-tex">j^{th}< / annotation>< / semantics>< / com:Math> transmitting antenna and the <com:Math><semantics>ith<annotation encoding="application / x-tex">i^{th}< / annotation>< / semantics>< / com:Math> receiving antenna and <com:Math><semantics>w→i<annotation encoding="application / x-tex">\vec{w}_i< / annotation>< / semantics>< / com:Math> is the noise at the <com:Math><semantics>ith<annotation encoding="application / x-tex">i^{th}< / annotation>< / semantics>< / com:Math> receiving antenna.

Claims

1. A method for transmitting frames of information (101, 201, 301) across a communications channel (106, 306, 406), the method comprising; a first conversion operation (102, 202, 302) for converting the frames of information (101, 201, 301) into a discrete-time Time-Limited signal (103, 203, 303) where the frames of information (101, 201, 301) are contained in a plurality of Finite Access Time (FAT) Degrees Of Freedom (DOF) (1307), meaning DOF for which the time it takes to access any of said DOF is finite; a second conversion operation (502) for converting the discrete-time Time-Limited signal (103, 203, 303) into a continuous-time signal; and a transmission operation (104, 204, 304) for transmitting the continuous-time signal (103, 203, 303) across the communications channel (106, 305, 406); characterized in that: the first conversion operation (102, 202, 302) uses a Block Toeplitz matrix and the plurality of FAT DOF is selected (1303) per a Mask-Matched Time-Limited FAT DOF (MTF) Design Step I (1302), in which for a fixed T s ≤ T s , m a x T_s \leq T_{s,max} (1301), a number (N) (1303), of FAT DOF is selected as N ≜ N S + N I N \triangleq N_S + N_I , where: N 𝒮 N_{\mathcal{S}} is defined as a number of Shaping FAT (S-FAT) DOF selected such that N 𝒮 ≥ N m i n N_{\mathcal{S}} \geq N_{min} in order to comply with a bandwidth (BW) constraint; and N 7 N_7 is defined as a number of Interpolating FAT (I-FAT) DOF obtained through creation of interpolated sampled frequencies inside an existing occupied band.

2. An apparatus for transmitting frames of information (101, 201, 301) across a communications channel (106, 306, 406), the apparatus comprising a first converter for converting the frames of information (101, 201, 301) into a discrete- time Time-Limited signal (103, 203, 303) where the frames of information (101, 201, 301) are contained in a plurality of FAT DOF (1303); a second converter (502) for converting the discrete-time Time-Limited signal (103, 203, 303) into a continuous-time signal (105, 205, 305); and a transmitter (104, 204, 304) for transmitting the continuous-time signal (103, 203, 303) across the communications channel (106, 306, 406); characterized in that: the first converter (102, 202, 302) uses a Block Toeplitz matrix (1104) and the plurality of FAT DOF is selected (1303) per a Mask-Matched Time-Limited FAT DOF (MTF) Design Step I (1302), in which for a fixed T s ≤ T s , m a x T_s \le T_{s,max} (1301), a number (N) (1303), of FAT DOF is selected as N ≜ N 𝒮 + N ℐ N \triangleq N_{\mathcal{S}} + N_{\mathcal{I}} , where:. N 𝒮 N_{\mathcal{S}} is defined as a number of Shaping FAT (S-FAT) DOF selected such that N 𝒮 ≥ N m i n N_{\mathcal{S}} \geq N_{min} in order to comply with a bandwidth (BW) constraint; and N ℐ N_{\mathcal{I}} is defined as a number of Interpolating FAT (I-FAT) DOF obtained through creation of interpolated sampled frequencies inside an existing occupied band.