Software Optimization for Flexible Wireless Networking Systems

Abstract

A method for configuring a flexible wireless networking system is disclosed. The system may include multiple Wi-Fi transceivers coupled in parallel with each other and selectively coupleable to one or more antennas by respective switches. The method may include obtaining topology information indicating whether and how each antenna reaches other antennas; determining a configuration change, including changing which and how many of the antennas are coupled to the Wi-Fi transceivers, based on the topology information; and initiating activation of a switch to couple one of the antennas to one of the Wi-Fi transceivers or deactivating a switch to decouple one of the antennas from one of the Wi-Fi transceivers. The configuration change may be determined using a machine learning model. The configuration change may include deactivating a currently active antenna and activating at least one alternative antenna having a different coverage area.

Description

[Technical field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims the benefit of U.S. Provisional Patent Application No. 62 / 931,112, entitled "Software Optimization for Flexible Wireless," filed November 5, 2019, which is incorporated by reference in its entirety herein.

[0002] TECHNICAL FIELD This disclosure relates generally to the field of configuring, optimizing, and operating flexible wireless networking systems. [Background technology]

[0003] Today, the existing techniques for building and deploying Wi-Fi networks involve installing a large number of individual wireless access points and connecting them to a switch with individual wired networking and power connections, usually Ethernet with Power over Ethernet to power the access points. The access points contain a small number of radios and are ideally distributed throughout the area they cover (usually an office building, etc.) to ensure sufficient wireless coverage to all commonly occupied areas of the deployment while minimizing the number of access points. Since these two goals, sufficient coverage and a minimal number of access points, are inherently contradictory, installers try to find a suitable compromise for their installations and often have to go back and update the location or number of access points to fill holes or provide density.

[0004] Every step of the process is expensive. Enterprise-grade access points are expensive, and even if they are given at a discount from a list price that is often over $1000, say, they sell for $100 each, the cost certainly does not allow an installer to double or triple the number of deployments, for example, because coverage holes are simply not possible. The labor required to run cables directly from the wiring closet to each access point location is expensive. Climbing a ladder and mounting the access points to the wall or ceiling (usually with special mounting brackets) is expensive. The power-over-Ethernet switch ports in the wiring closet to accommodate the wires running from the access points are expensive. The wiring in the wiring closet is expensive. The licenses required to operate the access points are expensive. Any controller appliances that manage the access points are expensive.

[0005] That is, the labor and material costs required to deploy a single access point ensures that customers must track and keep track of each access point, forcing customers to keep the number of deployed access points rational or reduced.Furthermore, even if customers wanted to flood an area with an excessive number of access points, this would go against manufacturer recommendations (express or implied) as radio contention and interference would cause the network to operate inefficiently and would overwhelm the algorithms in place today to regulate the network. Summary of the Invention [Problem to be solved by the invention]

[0006] One problem with traditional wireless installations is that each access point is discrete, tracked, named, and a valuable asset. Some attempts have been made to alleviate the pain. For example, access points have multiple ports and can pass some power from one port to the other. Thus, installers can daisy-chain access points together to avoid having to run wires into the wiring closet for each one. However, access points are not designed to support more than one or two daisy-chains, and the access points themselves lose bandwidth and power over the wired network.

[0007] Another problem with conventional wireless installations is that the radios are housed in discrete, self-contained, and expensive devices. This creates an inevitable tension between deploying enough devices for adequate service and not buying too many. Furthermore, each discrete box is a highly imperfect wireless device, given that it must hope for an antenna in a small enclosure with a good enough pattern to get through the obstacles that may be between it and the device it is calling, which may be far away. Thus, installation requires careful planning to ensure adequate visibility in all critical directions. [Means for solving the problem]

[0008] The appended claims may serve as a summary of the invention.

[0009] In accordance with some embodiments, methods and systems are disclosed for building and supplying wireless networking systems. Some embodiments use potentially long strips or planes of material with embedded wireless transceivers, power, and networking. Some embodiments use potentially long strips or planes of material with embedded antennas that are often switchable, thus providing far more opportunities for the transceivers to select the physical point of transmission and reception. Also disclosed are methods for dynamically configuring and reconfiguring this highly flexible antenna architecture using intelligent software processes to provide optimized coverage patterns for different deployments and under different conditions. [Brief description of the drawings]

[0010] [Figure 1] FIG. 1 illustrates one embodiment of the present invention showing a wireless transceiver and embedded antenna within a light bulb. [Diagram 2] FIG. 1 is an embodiment of the present invention showing a block diagram of a lighting embedded wireless access point. [Diagram 3] FIG. 1 illustrates an embodiment of the present invention showing an electronically steerable antenna. [Figure 4] FIG. 1 is a diagram of one embodiment of the present invention illustrating a set of electronically selectable antenna options. [Diagram 5] FIG. 1 is a diagram of one embodiment of the present invention showing an antenna strip sheet attached to a wireless transceiver. [Figure 6] FIG. 1 illustrates one embodiment of the present invention showing multiple transmission lines terminating in radiating antennas in both end-span and mid-span positions on a strip. [Figure 7] FIG. 1 illustrates one embodiment of the present invention showing a repeating pattern of antennas on an antenna strip with multiple transmission lines. [Figure 8] FIG. 1 is a diagram of one embodiment of the present invention showing common antenna transmission lanes and data lines for a selectable antenna switch to select between those antennas. [Figure 9] FIG. 1 illustrates one embodiment of the present invention showing a wireless transceiver and switch controller driving a selectable antenna switch on an antenna strip. [Figure 10] FIG. 1 is a diagram of one embodiment of the present invention showing a wireless baseband transceiver and switch controller driving multiple selectable baseband-to-broadband modulators with multiple antennas, such as on a strip. [Figure 11] FIG. 1 is a diagram of one embodiment of the present invention showing a headend controlling multiple regions of multiple antenna sets on a strip. [Figure 12] FIG. 1 is a diagram of one embodiment of the present invention showing a headend providing power and networking to a strip with a repetitive wireless system. [Figure 13] FIG. 1 is a diagram of one embodiment of the present invention showing a headend providing power and networking to strips with iterative radio systems each driving a iterative selectable antenna. [Figure 14] FIG. 1 is a diagram of one embodiment of the present invention showing the configuration of multiple selectable antenna sets of an antenna strip driven by a headend with a repetitive radio system embedded on the strip. [Figure 15] FIG. 13 is a diagram of one embodiment of the present invention showing resource allocation driven by the headend to select active units from the universe of units on one or more strips. [Figure 16] FIG. 1 is a diagram of one embodiment of the present invention showing a headend driving multiple wireless systems over a cable as opposed to a strip. [Figure 17] FIG. 1 is a diagram of one embodiment of the present invention showing a repetitive radio system powered by a string using AC power. [Figure 18]FIG. 13 illustrates a possible advantage of an embodiment of the present invention showing that access point transmission locations may be electronically set to points among a higher density of selectable locations, thus more closely emulating a situation where access points may be physically relocated from their current locations to their optimal locations. [Figure 19] FIG. 1 is a diagram of one embodiment of the present invention illustrating resource management and topology discovery across multiple headends or discrete access points. [Figure 20] FIG. 1 illustrates a method for installing multiple embodiments of the present invention. [Figure 21] FIG. 1 illustrates an example deployment of a flexible wireless networking system. [Figure 22] 22 illustrates a scenario in which three wireless client devices are within range of the strip shown in FIG. 21. [Figure 23A] ~ [Figure 23B] FIG. 2 illustrates an exemplary embodiment of an antenna patch. [Figure 24] FIG. 2 illustrates an example antenna radiation pattern for an entire building taking into account channel reuse. [Diagram 25] FIG. 25 is a diagram showing an example of a result of pruning on the antenna radiation pattern shown in FIG. 24. [Figure 26A] ~ [Figure 26B] FIG. 1 illustrates an example of pruning results for an antenna radiation pattern with overlapping circles. [Figure 27] FIG. 1 illustrates an exemplary embodiment of a system and method for software optimization of a flexible wireless networking system. [Figure 28] FIG. 1 illustrates an exemplary embodiment in which multiple antennas are used for transmission and reception. [Figure 29A] ~ [Figure 29B]1 illustrates two similar transmissions being picked up sequentially by antennas in respective configurations having two different overlapping antenna sets according to some embodiments. [Diagram 30] FIG. 1 illustrates an example embodiment of a method for dynamically configuring a flexible wireless networking system using machine learning. [Diagram 31] FIG. 1 illustrates an exemplary embodiment in which the antenna is divided into a specific repeating pattern. [Diagram 32] FIG. 1 illustrates an example embodiment of a strip having a repeating pattern over the length of the strip, including between two radios. [Figure 33A] ~ [Figure 33C] 1A-1C illustrate different possibilities for antenna selection over distance for an exemplary strip according to some embodiments. [Figure 34A] ~ [Figure 34C] FIG. 1 illustrates an example embodiment including individual switchable antennas along a strip attached to a single radio, with wireless client devices moving relative to and along the strip. [Figure 35A] ~ [Figure 35C] FIG. 1 illustrates an example handoff scenario using a limited set of radiators. [Diagram 36] FIG. 1 illustrates an example embodiment of a system and method for software optimization of a flexible wireless networking system including a handoff manager. [Figure 37] FIG. 1 illustrates an exemplary embodiment in which a handoff manager is integrated into a configuration manager. [Figure 38A] ~ [Figure 38C] FIG. 1 illustrates an exemplary embodiment in which a strict mapping of antenna elements is maintained following a handoff. [Figure 39A] ~ [Figure 39B]1A-1C illustrate the use of three-dimensional antenna shapes on strips to generate directionality, according to some embodiments. [Diagram 40] FIG. 1 illustrates an exemplary embodiment of a four-way switch fabric in which no radio can be disconnected from the antenna. [Diagram 41] FIG. 1 illustrates an example embodiment of a full matrix switch of single-pole, single-throw switches to which any combination of RF chains and antennas can be interfaced. [Diagram 42] FIG. 1 illustrates a computer system upon which an embodiment may be implemented. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] One theme that connects many of the disclosed invention embodiments is that they leverage methods to increase the density of available wireless transmission locations, thus potentially enabling electronic reconfiguration of the physical layout of the wireless network without the need for human intervention or physical rearrangement of actual discrete assets. A possible advantage of the disclosed form factor is that it can be easily incorporated into architectural or cabling plans and installed by electricians rather than skilled networking labor.

[0012] Figure 1 shows the general architecture of a wireless light bulb 110. A Wi-Fi transceiver 120 is presented. Usually this is available as a complete system on a chip, meaning it includes a CPU, memory, and operating system along with its radio and media access control (MAC) layers. Being a light bulb, it also has LEDs and a sufficient LED controller attached to the system on chip (SoC) so that the SoC can dim the LEDs (usually by pulse width modulation) based on network demand. No lighting components are shown in the figure, since the embodiment is not just about light bulbs. The Wi-Fi transceiver 120 acts as an access point (AP) and relays information from the network 110.

[0013] FIG. 2 shows an elaboration of the embodiment of FIG. 1. The network 200 is implemented using networking over powerline technology (which typically does not need to be high capacity) routed over a mains power source 210. The light bulb has a power source 220 with enough capacity to drive its lighting and wireless needs. A powerline network transceiver 230 connects a CPU / controller 250 to the powerline network 200. A Wi-Fi transceiver 290 provides generic Wi-Fi functionality. Such transceivers are typically structured such that the CPU 250 handles the higher layer networking configuration, and the transceiver itself is programmed by the CPU to adopt the correct wireless mode, including how many Base Service Set Identifiers (BSSIDs) to offer, what their addresses are (often via base addresses and masks), what wireless settings are required, etc. Typically, the software on the CPU has an access point manager 260, which handles the associations and state machines for each client, and also handles the list of clients 280, and the BSSID and Service Set Identifier (SSID) offerings 270. One or more antennas 295 are provided.

[0014] The illustration of the light bulb as screw-in in FIG. 1 is merely one embodiment, as light bulbs come in many form factors including plug-in, snap-on, and fluorescent tube bulbs (even though the alternative lighting technology is LED). The power and power line networking technology in some embodiments operates via a fluorescent ballast. Other embodiments are fitted with a standard two or three prong outlet plug and therefore plug into a power outlet, and further embodiments employ a night light configuration. Other embodiments wire directly to a power source, such as via a wing nut in a junction box.

[0015] In other embodiments, powerline networking elements are not used or even present and Wi-Fi transceivers are used for mesh networking.

[0016] The light bulb as a literal form factor is not the only possible form factor. There are other places where an access point can be introduced. By simply removing the lighting component from a light bulb embodiment, one can have a "night light" that is simply a radio with no lighting. Alternative embodiments are different, such as those built directly into the outlet, those hardwired in place of a standard outlet, controllable smart outlets, or GFCI outlets, each carrying the necessary features of a simple plug that is controllable and has GFCI protection.

[0017] Built-in radios such as those according to the above embodiments can be installed such that the antennas are not configured by the installer. Antenna patterns are often important, especially when the device is placed against a ceiling or wall or furniture. Having only one fixed set of antennas would then be problematic if the resulting antenna pattern relative to the environment results in an overall distorted cell that fails to provide adequate coverage.

[0018] A series of embodiments solves this problem by using an automatically movable antenna. Figure 3 shows two embodiments. Figure 3a shows an antenna 300 with one axis of freedom attached to an electrically movable device 310 such as a servo, motor, or electrically deformable material. Figure 3b shows two axes of freedom by attaching one movable device 330 to another movable device 340 in different orientations such as a cross. Omni-directional antennas for Wi-Fi only need to be 1-2 cm long, and the required mechanical servos and motors can now be manufactured at very low cost and in small size. Thus, multiple movable antenna packages can be easily placed in a small form factor such as a light bulb. In other embodiments, instead of servos and motors, springs and variable length struts are used to position the antenna, and the variable length struts can be memory wires, rack and pinion systems, solenoids, voltage sensitive materials, or other similar mechanisms. The choice of material is a manufacturing decision, and there is a tradeoff between power required for operation, cost, and complexity, and such a choice is obvious and does not require experimentation or further invention. In some embodiments, multiple antennas are mounted together on a single movement mechanism.

[0019] If mechanical actuation is not desired, multiple antenna patterns can be switched. Figure 4 shows one embodiment of antenna switching. Multiple antennas are provided in different directions, positions, or different shapes and patterns. In this figure, each antenna 400 is provided with its own one-way switch 410, or alternatively, a reduced number of multi-way switches can be provided. The antennas may be arranged in different directions, in two-dimensional or three-dimensional space. The antennas may be separate wires or printed on plastic or solid circuit boards, such as microstrip antennas. This can often be done very cheaply. Some embodiments provide enough antenna possibilities for each antenna with a selection so that the system can search through various antenna patterns and find the best antenna pattern to use. Long ago, in the early days of Wi-Fi, per-packet antenna switching was used, where the transceiver had enough time to compare the signal strength of multiple antennas during a disposable training preamble (the circuitry was usually not sophisticated enough to measure much else), and the system could then select which antenna it wanted to receive the main signal from. With the speeds increased with OFDM, this method became unusable, as there was simply not enough time to detect. Also, in MIMO, multiple antennas transmit different signals, so it becomes meaningless. Still, that doesn't mean that the location of the antennas doesn't matter. Rather, today's access points have antennas tacked prominently to the housing in reasonably orthogonal locations. (External antennas are somewhat rare, as they tend to fall off over time, changing the access point's pattern.) But when access points are placed in physically constrained locations that prevent a priori placement, such as light bulb sockets or junction boxes, the need returns for antennas that can be remotely and automatically adjusted, not per packet, but per placement, and even over time based on moving traffic patterns, etc.

[0020] Access points are traditionally placed with great care and precision by both planners and installers. Ideally, one would rather place them in any convenient location and run antennas remotely in the best location. However, with MIMO, this is difficult to achieve with cabling, and each antenna needs to be placed at a certain distance away, with enough independence between them to capture independent spatial channels at the same frequency. Few people are willing to pay the expense of running 2-4 coaxial cables just to dangle an external antenna on the end of them, coaxial cable crimping tools are not a common feature in an installer's tool belt, and usually, if one can run a coaxial cable and install an antenna, one can save money and effort by installing an access point in the same place instead. In fact, for these reasons, most access points sold today only contain internal antennas.

[0021] However, this installation problem can be mitigated by using cheap, possibly flexible, and easy to install sheets of material with embedded antennas. Figure 5 conceptually illustrates an antenna strip sheet. A Wi-Fi transceiver 500 connects to a strip 520 via one or more lines 510 for one or more antennas. The strip sheet is then positioned across space, thus offering the possibility of multiple antenna locations.

[0022] In some embodiments, the strip sheet is a printed product, usually a plastic flexible circuit board, very similar to cheap LED lighting strips. In some embodiments, the strip is made of repeating units of addressable antenna sets with supporting infrastructure. The antenna sheet, in some embodiments, is attached to a head-end unit and, in some embodiments, addressably drives various components from passive antennas to active transceivers. The goal is that these sheets come in long spools that can be inexpensively put in and cut as needed with little financial pain, thus allowing for unskilled installation and minimal waste costs.

[0023] In some simple embodiments, the strip sheet has one set of antennas, as shown in FIG. 6. In some embodiments, the lanes 640 are surrounded by sufficient printed shielding lanes on the same plane 620 to prevent radiation or absorption from the lanes themselves, and laminated shielding on the top 610 and / or bottom 650 printed planes to transmit the signals of each antenna 630, which in some embodiments are grounded and separated from the lanes by an insulating space or another insulating layer in the same plane. In other embodiments, other shielding mechanisms are used. With shielding as shown in the figure, most of the non-radiating portion of the strip acts like a cheaply made coaxial cable, but since there is no coaxial element or cable, it can be manufactured using the same methods as an LED strip. One possible advantage is that, like coaxial, the signal attenuation can be kept approximately linear with the length of the antenna strip. The selection of the width of the shielding strip can be easily done by one familiar with the art of microstrip transmission lines without undue experimentation and can be selected based on the desired quality at the desired price. At the radiating location, the shielding is omitted and appropriate radiating and impedance matching structures are used. In FIG. 6a, the four antennas are shown side-by-side at the unshielded end of the strip, and while exemplary, such adjacent aligned antenna options may be performed, such as arranging parallel PIFA antennas with their primary radiating structures in the direction of the figure. However, in some embodiments, the antennas are arranged in different directions, patterns, and different shield and ground plane configurations. For example, FIG. 6b shows two sets of orthogonal radiating antennas surrounded by shielding material on the lane plane. (Again, although not shown in detail, the PIFA antennas may be printed in such a primary direction). Other embodiments use different lane plane shields, or omit the lane plane shield in the radiating elements, as in part a of the figure. Other embodiments remove a layer of shielding (e.g., the top) above the radiating plane, but leave the bottom, to make the antenna radiate in a preferred direction perpendicular to the plane of the strip, as is typically done with patch antennas.The length of the radiating element that can be accommodated is as long as the long side of the strip allows, but practical considerations such as wanting to create a reasonable beam pattern may constrain the length, array pattern, and potential gain to a few dBi or less. The shape of the antenna is also a choice based on need. In some embodiments, there are multiple radiating elements attached to the lane, some embodiments do this to allow for different pattern representations, other embodiments do this to allow for different frequency responses, such as having multiple lengths for optimal resonance based on the frequency band used.

[0024] In further embodiments, multiple antenna sets are placed on the same strip, but at different locations. A simple embodiment is shown in FIG. 7. In this particular embodiment, the radiation pattern is repeated at least once down the strip. Each antenna 720 maintains its own lane 710 back to the head of the strip, within the surrounding shield 700. In this and the following figures, holes in the other shielding layers are not shown, but it is clear from the previous discussion that in many embodiments there are holes above at least the radiating portion of the antenna to allow the signal to escape in one or both normal directions to the plane, and as before, the shielding around the radiating antennas on the same layer may be different. In this particular figure, the lanes for the second antenna set are shown on the same layer as the first antenna set. In other embodiments, the antenna lanes are distributed on different layers, all with appropriate shielding between them (one particular embodiment zigzags the lanes vertically between each layer, such that the lanes on one layer are surrounded by the shielding between the lanes on the layer above and below), and then it is possible to connect to the higher radiating layer with the antennas exposed to the outside world with vias.

[0025] FIG. 8 shows further embodiments using switched antennas. In these embodiments, antennas 840 create a repeating set along the strip as many times as necessary, with one lane 810 dedicated to each antenna. The strip is fitted with active switches 820. The power lanes and ground return for the active switches are not shown in this diagram as they can be clearly wired by those skilled in the art. However, the data lines 830 for the switches are shown. In some embodiments, this is a serial line that communicates to the antenna switch / controller 820 that it is the switch that should be active. In one embodiment, each switch has its own dedicated switch control line. Another embodiment uses a simple pulse code to chain the switches / controllers together, with a master controller for the data lines near the Wi-Fi transceiver that emits a series of rectangular pulses of a specific address length followed by trailer pulses of different lengths. Each switch / controller in its embodiment electrically terminates the line, counts the incoming pulses, does not emit the first pulse, and emits the second and subsequent pulses, including the trailer pulse, downstream to the next switch. If only one pulse comes in before the trailer, the switch closes and connects that antenna to the antenna lane, otherwise the switch opens and disconnects the antenna. Thus, the master controller emits as many pulses as the count of the switch it wants to activate, starting from 1, followed by the trailer. For example, sending a pulse of 6 counts and a trailer will open all switches on the strip except the 6th switch counting from the head end. Obviously, other codes may be used, including binary addressable codes using a variety of different encodings, including variable pulse width, Manchester, and other encodings known in the art. The particular embodiment just described is merely a simple code that can be implemented with minimal circuitry. Note that shields are not shown in this diagram for clarity, as we have already discussed how they should be placed.For inter-plane shielding, now that we have mounting components, some embodiments leave holes in the shield layer for the solder pads of the active components on the layer below. Another approach is to move the antenna lanes to a different layer below the top mounting layer and bring the lanes to the surface through vias, and in one particular embodiment, vias are placed on the solder pads of the switches, which shield the direct radiation with their own body, preventing accidental leakage. A further embodiment is to use the top shield layer, directly below the top mounting and radiating layer, as a ground layer for the active components, either occasionally connecting the ground through vias to their own ground line printed on the top layer, or always grounding to the second layer through vias.

[0026] FIG. 9 further illustrates a master controller 910 with switch lines 930, such as those shown as S0-4 (or as desired), and a Wi-Fi transceiver 900 with antenna lines 920, such as those shown as A0-4 (or as desired), connected to a strip or similar structure 940. In other embodiments, a different number of antennas may be used, and the pattern and distance of the antenna repeats are manufacturing options. Ground / shield and power connections are not shown, as they are self-explanatory. Note again that other embodiments use multiple antenna switches that switch multiple antenna lanes from the same control line. Some embodiments use one switch controller per antenna set, but drive multiple single antenna switches from one switch line coming from the controller.

[0027] The antenna strip has power losses all the way through, and because it is an analog process, the transceiver needs to do some work. In some embodiments, this is accommodated by bi-directional amplifiers mounted between the transceiver and the strip, or placed at one or more points along the strip. In one embodiment, the amplifiers are at the head end of the strip. In another embodiment, the amplifiers are at one or more locations along the lanes, thus splitting the lanes at their insertion points. In another embodiment, the amplifiers are mounted between the lanes and the switches, one per switch, thus not splitting the lanes. Other arrangements are possible.

[0028] However, for very long runs, or where radiative concerns are paramount, or for reasons of cost or ease of installation, other options exist. FIG. 10 shows an embodiment using a baseband antenna strip 1030, shown diagrammatically for clarity, since the method of constructing the strip has already been disclosed. The pre-RF Wi-Fi transceiver is now reduced to a baseband transceiver 1020. Instead of RF antenna lines, there is at least a pair of baseband analog lines I and Q 1025 for the actual complex part of the ultimate wideband signal. At least one bidirectional baseband to broadband converter 1050 (two repeat units shown in the figure) modulates and upconverts baseband to broadband on transmit, and downconverts broadband to baseband on receive. The Ch line or lines specify the channel and transmission characteristics. In some embodiments, the Ch line is just one line, carrying the channel carrier signal, and the converter detects transmission or reception according to the state of the IQ line. In other embodiments, the Ch line carries a signal describing the channel to be used, and the converter has a resonator inside it to generate the required frequency, and the channel description signal in certain embodiments consists of a low frequency signal, such as a crystal driven, and a multiplication signal that specifies the relative frequency to be generated from the crystal line. In other embodiments, the Ch line also carries the Tx / Rx line, and the converter does not decide whether to transmit or receive by sensing the IQ line, but instead uses the Tx / Rx line (usually driven by the MAC via the transceiver). As shown in the figure, some embodiments select which antenna set to use by selecting which converter to drive. In some embodiments, the converter does not use a controlled switch, such as the antenna switch described above using a serial line, but instead does not power up unless switched on using a power switch on the power line. In other embodiments, the converter has an enable line and is driven by a controller, such as a serial controller as used for the switch above.One possible advantage of these embodiments is that the baseband signal is still conveyed with less error, and the converters are small and cheap enough that they can be attached to the antenna strip in an almost disposable manner.

[0029] Further embodiments can be described that use a low frequency real signal carried on the channel and an additional conversion stage instead of using a full baseband converter. This two stage conversion requires replacing the IQ lines with low passband lines, usually one instead of two per antenna, set at a low enough, non-radiating frequency to allow the signal to be represented completely as a real voltage. A similar setup to the previous diagram would result. A possible advantage of using an intermediate passband is that it reduces the number of lines and the complexity of the converter.

[0030] In all cases, multiple control lines and antenna lanes can be implemented within a strip (whether an antenna strip as disclosed above or a transceiver strip as disclosed below), thus making it possible to produce embodiments with strips that can transmit or receive on more than one antenna set at the same time. FIG. 11 shows a multi-region strip 1120. In some embodiments, the regions are contiguous, and the strip repeats a new region at intervals. In some embodiments, one antenna set 130 per region is available by count, so if there are five regions and five command lines, the first command line can choose and control which one antenna set in the first region to use, etc. In other embodiments, the regions are one-of-n, meaning that within a repeat region interval, the first antenna set belongs to a first selected set, the second antenna set belongs to a second selected set, and the first control line selects which region's first antenna set to use, etc. (Note that some of these embodiments are interleaved forms of the contiguous previous embodiments).

[0031] In many cases, an antenna strip will suffice. However, in some cases, analog losses remain unacceptable and a purely digital transmission mechanism is required to extend the length even further. The good news is that the full Wi-Fi SoC itself is cheap enough to be packaged in multiples on a strip.

[0032] FIG. 12 shows a transceiver strip 1230. At a minimum, the Wi-Fi system 1240 includes digital-to-analog transceivers. However, in some embodiments, the Wi-Fi system includes an entire system-on-chip. In these embodiments, a network line connects the systems together. In some embodiments, the network line is a high-speed bus. In some embodiments, the network line is a tappable network, such as hubable or modified point-to-point Ethernet. In some embodiments, the network line acts as an addressable backplane. In other cases, the network line is a broadcast medium. In some embodiments, the headend 1200 provides the networking 1220 and power 1210 for the strip. One possible advantage of these transceiver embodiments is that multiple systems can communicate simultaneously, on the same or different channels, although limited by power and bandwidth. In this way, these strips can be made to act very similarly to inexpensive LED strips for the installer: measure as much as you need, route where you want, and cut if desired. The system can then determine how much strip to use, when and where to use it, as disclosed below.

[0033] Since the strip offers a variety of possibilities, the previous embodiment with multiple antenna patterns per bulb form factor can be used directly on the strip as shown in FIG. 13. For clarity, only one antenna lane and its switchable options are shown, but this combination is generic and can be extended to multiple lanes and options. The switches 1340 and antennas 1350 can be operated in the same manner as those in FIG. 4. FIG. 14 shows an embodiment with a generic combination, where the strip 1430 is divided into repeating meta-areas including a Wi-Fi system 1440 acting as its own controller and a head end of sub-areas 1450 (two shown, but any number is possible), each of which includes an antenna set 1460 (four shown, but any number is possible), which may include any of the disclosed embodiments, such as passive multiple antenna lanes, active switch lanes, transformer lanes, etc.

[0034] The strips above have been described as flat, two-dimensional objects. However, some embodiments are three-dimensional, as antennas may need to be three-dimensional. In some embodiments, the antenna set is attached to a daughter strip. In some embodiments, the antennas are soldered or hard wire connected in place. In some embodiments, the strip itself is crimped, folded, or rotated at intervals to allow it to cover different orientations before returning to its original orientation. In some embodiments, the strip has mechanical connectors, such as rotatable connectors (some free to rotate with brush rings, or constrained with cables) or foldable or flexible bridges, at repeating intervals to allow it to change orientation when installed. In some embodiments, the antennas are motorized, as disclosed above. Some embodiments fit together with some parts being flat and others having these options, according to manufacturing specifications.

[0035] The strips may have an efficient form factor. However, the interconnection techniques that have been disclosed on the strips may also be performed outside of the strips. Thus, other embodiments use the structures disclosed above, but separate the circuits using wires or cables. For example, some such embodiments mount the antenna sets on their own circuit boards (flexible or non-flexible) and connect them by cables that include the same structures. Other embodiments simply separate the repeating units into separate strips or circuits and connect them with cables. This may be done in the field in the same manner that LCD strips are cut and cabled in the field (including using solderless contact closed cables) or may be done in this manner at the time of manufacture. The cables may or may not have plugs to allow for field reassembly of strings to desired lengths.

[0036] Because cabled strings have a useful form factor, a further set of embodiments is directed to powering the strings with a limited power budget, such as from one or more Power over Ethernet (PoE) drops.

[0037] The bulb and strip embodiments converge to a particular set of embodiments. Long bulbs may require more than one antenna set, and thus in some embodiments the bulb includes two or more antenna sets, as does the strip, or installed by wire. Further embodiments, not surprisingly, allow the strip to incorporate lighting and also operate as an LED strip.

[0038] It should be noted that the number of antennas, the length of the strip, and the layout and spacing of antenna sets and repeating units or regions on the strip are manufacturing options and are not intended to be limited by the figures or description.

[0039] Existing buildings, and by default new buildings, often have PoE cables in place for rooms. These PoE cables are installed today for traditional Wi-Fi deployments, where one access point is expected to connect to the PoE port and thus receive its networking and power. Unlike traditional AC power, PoE has significant wattage limitations, so it is highly unlikely that a strip containing, for example, 20 active Wi-Fi transmitters at full power could be powered from a single PoE port. Beyond power, other resources such as wireless interference, network bandwidth, and even radiated power human exposure concerns can lead to the need to dynamically budget.

[0040] Some embodiments address the need to stay within a power budget by limiting the number of activated circuit components. (Without loss of generality for other resource constraints, the term "power" may be used in place of the general resource in the following.) FIG. 15 illustrates some embodiments that perform resource budgeting. For one or more strips 1530 (or strings, if cabled) with multiple Wi-Fi systems, including a complete system on a chip, the resource budgeter 1520 may request that the SoC power down to a power-saving state. Some embodiments allow the SoC to wake up again using Wake-on-LAN on its wired network. Some embodiments use a specific addressable power controller switch to directly control the power of the SoC. The resource budgeter may also request that a set of multiple transceivers turn down or off individual transceivers.

[0041] In some embodiments, the resource budgeter can dynamically measure the available power, in other embodiments it is manual. In some embodiments, the resource budgeter can measure the length and contents of the string or strip using a discovery protocol (digital, broadcast, or using keyed passive components such as resistors or capacitors at different values ​​corresponding to different cut points on the length of the string, etc.). (Assuming a 100 kΩ resistor spans each cuttable region across two lanes of the string, and cutting the string to include five regions, there will be five parallel 100 kΩ resistors for a total of 20 kΩ.) The resource budgeter in some embodiments is located at the headend 1510. In other embodiments, the resource budgeter is located at the repeating unit of the strip or string.

[0042] Note that one or more PoE ports may be combined to increase the available power. Below are several embodiments of resource-managed PoE strip sets.

[0043] FIG. 16 shows an embodiment similar to the previous figures, but with wires or cables 1650 connecting the individual units 1660 together as a string, back to a headend 1610 powered by one or more PoE lines 1600. In this embodiment, the entire unit may be thought of as a cheap Christmas light string, except that the service provided is wireless. Also, as before, other embodiments replace PoE with other networking and power options. In some embodiments, as an option, or the only option, an AC power source is used to power the headend, and it connects to the network by other means, such as an Ethernet port. This is a slight modification of using PoE for power by allowing the use of a local power source. Since the cables can be long, it may be more convenient to provide power locally than in traditional wireless PoE applications. As before, the cables may be permanently wired or may be pluggable. Some embodiments use only power lines, replacing PoE with power lines and power line transceivers for the headend. Some further embodiments then provide low power and separate network lines to each unit on the string. Another further embodiment keeps the AC power on the line but has a separate wired network on the cable. Another further unit uses power line networking and power technology throughout the string.

[0044] FIG. 17 shows an embodiment that omits the head end entirely and uses power lines to power individual units on a string or strip by bringing a local power supply 1720 to each individual unit 1710. Some embodiments also use a power line transceiver 1730 to access wired networking using the same power. Some embodiments relocate resource budgeting to one or more individual units of the string, such as the first unit, other embodiments relocate resource budgeting to a different unit or a different head end, or outside of the string, such as a server, appliance, or cloud, and other embodiments omit resource budgeting entirely. In these embodiments, each unit is equipped with a power line transceiver to communicate over AC power. Some embodiments use junction box wiring instead of plugs.

[0045] When a building is filled with Wi-Fi (or other protocols such as Bluetooth or 4G / 5G) light bulbs or strips, perhaps densely over-provisioned, a control problem arises. It doesn't make sense to have each radio fully powered up and transmitting, as doing so would make interference problems more severe. However, the flexibility to choose which radios to use among the total set of radios allows us to exploit the richness of the deployed location options to approximate the ever-changing optimal location. This approximation can arbitrarily approach global optimality with radios that can be driven to the right places on some kind of mobile platform (e.g., a robot).

[0046] Figure 18 compares in part a an ideal situation where base stations 1800 and 1830 are physically moved to optimal locations 1810 and 1840 in response to changing network needs, with part b, where an approximation to part b is obtained by densely deployed resources, most of which are disabled (small empty circles 1820), but some are powered on and operational as needed. Without the ability to control this and create a reasonable approximation of optimal or reasonably deployed resources, any form of dense deployment is doomed to failure, which is one reason why existing Wi-Fi light bulb manufacturers have not bothered to turn their light bulbs into access points and get into the Wi-Fi infrastructure business. (Another reason is that, as mentioned above, doing so was prohibitively expensive and impossible to install before this invention.)

[0047] FIG. 19 expands on the basics of dense or over-deployed resource management. Radio resource management today is mostly limited to standard deployments. In minor cases of over-deployment, some APs can be powered down, and in some cases some extreme APs can be presented or automatically turned off. However, these are designed for deployments where the excess resources that need to be shielded are much less than the number of active resources, whereas what is disclosed here is for the case where the excess resources far exceed the active resources, as in FIG. 18b. The scale matters. If properly deployed, a densely deployed network in a typical enterprise campus could have over 10,000 radio resources, in some cases approaching a density of more than one antenna set per square yard in the area of ​​the physical deployment. Existing radio resource management techniques have never been built for such density or scale, and are therefore limited to operating on expensive, carefully installed access points.

[0048] The location of the resource manager is flexible. In some embodiments, the manager is software located on the controller appliance. In other embodiments, the manager resides in one or more locations as software on a server, desktop, or in a headend or CPU located elsewhere within the invention. In some embodiments, the manager runs in the cloud. The resource manager in some embodiments exists as only one entity, while in other embodiments, multiple such resource managers, or elements thereof, may exist and coordinate with each other.

[0049] The topology discoverer 1930 is responsible for understanding just where all the resources ended up. This can be measured in radio space (measured by signal attenuation between two measurable objects), in physical space (measured by location, such as using GPS or BLE), or a combination of the two. Unlike traditional Wi-Fi networks, resources can be highly over-deployed, and in many cases there are many other resources in range of any given resource, making topology discovery very simple. However, some care must be taken when measuring the topology. In power- or resource-constrained systems, it may not be possible to power up each antenna set at once. However, bootstrapping mechanisms are available. One such mechanism is to record the topology by turning on a subset of resources (such as the largest subset available based on resource constraints), then turn off some smaller subset of resources, turn on another subset that was not yet turned on, and merge the two subsets together. This is trivial if the topology is managed in a graph format. If the topology is also laid out in some metric space, a best fit must be done, usually some kind of fit will have to be found, as there is usually statistical noise in the measurements. Due to the nature of topology measurements, they can often be done in an operational environment, and usually the available resources will be less than the maximum budgeted resources, but even if this is not the case, protocol methods (such as requesting a quiet period to go "off-channel" as in 802.11k) can be used to allow existing resources to be powered down and other resources to be turned on. Note that resources are at an antenna-like granularity, allowing the system to cycle through and search antenna patterns, not just transmitters, complete radios, complete access points, etc. Part of the topology should include the distribution of users and their wireless clients.

[0050] Once the topology is known, the resource manager 1940 can evaluate the dynamic network utilization and demands and determine what the new best distribution of resources should be. Some embodiments consider TCP load. Other embodiments consider application specific knowledge such as tracked application state. This is very useful for streaming applications where video streams can be predicted from their exchanges. Other embodiments use past statistical history. Other embodiments use specific resource demands. Some embodiments consider minimizing service interruption, or at least weigh the cost of interruption of service to existing clients, such that they also evaluate service interruption for new loads that cannot be adequately served by the existing configuration. Once the resource manager 1940 decides what changes to make, it communicates to the active resource allocation storage 1950, which stores what has been requested. If the resource changes are non-disruptive, such as changing the local antenna pattern, those requests are communicated directly to the headend of the strip and other APs to implement. However, if distances are involved, more work is required. The inter-antenna set handoff manager 1920 is responsible for determining how to perform the reconfiguration over distances.

[0051] Distance issues are important. For example, it may be desirable to move an AP so that it is in view of clients. This may result in some clients being out of range. In some cases, this is an acceptable risk and the handoff manager 1920 decides to do nothing about it and performs the move. In some cases, multiple clients are stranded and the blockage will require the client to relocate. In some embodiments, this happens by disassociating the client. In some embodiments, the client's access point requests a load balancing or handoff protocol exchange (such as in 802.11k / 802.11r) to indicate that the client wishes to move away from its current access point (and likely channel). These protocol requests may include suggestions or requests based on the protocol, as long as they indicate a good suggestion as to where the topology is headed.

[0052] One technique is to preserve the BSSID of the client even when the antenna set is changed. If the old and new antenna sets belong to the same medium access controller, this is trivial. However, if the old and new antenna sets belong to different MACs, the handoff must request the old MAC (and possibly the old AP) to power down and the new MAC to power up. In some embodiments, the new one is powered up first, taking into account resource budgets such as power. Once the new one is up, the handoff manager hands over enough access point state to allow seamless operation. This may include security tokens and counters, client state machines, application tracking state, etc. In some embodiments, this state is streamed from the old access point to the new access point until the access point has established state and is listening to its radio to ensure it is ready to start transmitting. In some embodiments, both the old and new radios are powered up and the new radio just waits for the new MAC to start taking care of it, which may be done at the MAC level or by tricks such as setting the BSSID mask to an invalid or useless value so that the set does not respond. Once the handoff is ready, the old and new APs then agree on the handoff and perform it in as atomic a manner as is reasonable. In some embodiments, the old and new APs agree on a specific cutover time. In other embodiments, one of the old and new APs waits for a message from the other (over the air or on the back network). In some embodiments, the handoff may occur only in the head unit properly, while in other embodiments, it may occur between any two MACs and their APs. Note that the client does not see the handoff, but experiences the AP magically jumping from one location to the other, even though in reality a transition has occurred between the two APs. This is because a smaller set of BSSIDs is assigned to a larger set of APs, breaking the traditional one BSSID per one AP mapping.This is entirely within the scope of the standard, since such transitions simply appear as physical relocations or failover events. Care must be taken not to violate client expectations without fair warning, for example by transitioning radio functions that are not permitted to be changed while the client is associated. The handoff manager is responsible for such care, and if it determines that it is necessary to violate its rules, it must assign new BSSIDs where possible to prevent client confusion. A channel change may not require a BSSID change, since access points change channels periodically in conventional systems.

[0053] Many of these embodiments have what could be a simpler installation. In many cases, an installer can simply walk to the beginning of a long hall, for example, unroll the strips, and perhaps simply place them on the ceiling tiles. They may also be painted to fit the strips in the visual space they are installed in. For temporary installations, the strips can simply be unwound when finished. For plug-ins or bulbs, the installer can simply screw in what is dense enough for the room. This density can eliminate the need to do a "site survey" or think too hard about interference and density effects. As long as enough items are installed, the system may have a good resource activation solution. For strips that do not include transceivers, it may even be possible to upgrade the radios from one standard to a different standard with different hardware by simply replacing the head end and leaving the strip in place.

[0054] FIG. 20 shows an example deployment using both strips and bulbs on a floor plan, thus illustrating the potential benefits of the disclosed invention. The long string 2000 is an implemented strip or string. The rectangles 2010 are embodiments of bulbs (large tube lights in this case). The strips are equipped with a head end powered by PoE (not shown) based on convenient drop locations. The bulbs run back via mains power to a power line network converter that bridges to an Ethernet backbone (also not shown). In the closet (not shown) is a resource manager that oversees the entire system.

[0055] As described in detail above, some embodiments use a strip (or wire or sheet or other distance or area filling material) with multiple switchable antennas disposed throughout, connected by a backplane, that are connected to multiple radios on the strip. In some embodiments, the various antennas may have different patterns and positions at multiple distances along the strip, and thus the disclosed system can select one or more antennas with their different patterns at different positions along the strip to potentially organize an overall pattern of nearly any shape based on a switching decision. The strip has many different antennas in many different patterns arranged to organize them into one or more cohesive patterns.

[0056] FIG. 21 shows an example deployment of a flexible wireless networking system as described above, including multiple radiating antennas 2110 coupled to antenna lines 2106 via respective switch / controller pairs 2104, and showing spatial antenna lobe patterns 2112 (for coverage) emanating from the strip 2100 into the space 2114. The shapes of the radiating antennas are shown for illustrative purposes. In other embodiments, different antenna shapes may be created (such as micropattern printing) to generate different radiation patterns, and combinations of two-dimensional or three-dimensional radiating material shapes may be used as described above. Also, the orientation of the strip 2100 relative to the space 2114 being radiated is variable. For example, the figure shows edge-on radiation for simplicity of explanation, and an edge-on pattern is the intent. However, other embodiments may implement face-on radiation or anything in between. Note that while FIG. 21 shows all the switches as switch / controller pairs 2104 on the data lines 2102 as one embodiment, other embodiments may include parallel switch lines dedicated to each switch and controlled separately as on / off signals. The system has the choice between any or all of these patterns, and software (or hardware) can change switch states to organize the individual patterns into larger patterns. The choice of pattern may depend on the environment and operational or performance goals.

[0057] In FIG. 22, there are three wireless client devices within the range of the strip 2100 shown in FIG. 21. Note that the distance between antenna pairs, such as the distance 2210 between the fourth and fifth antennas, may be any distance ranging from just a few inches to many yards, and may vary between different pairs of adjacent antennas on the strip 2100. In the illustrated example, if device A 2202 is the device that cares about the system obtaining the maximum SNR, it should include the left-most antenna in its selection. If it also cares about devices B 2204 and C 2206, it should include the middle and right-most antennas. However, if it wants to exclude signals from devices B 2204 and C 2206, it should exclude the middle and right-most antennas instead.

[0058] Note that this selection is quite different from antenna selection in the traditional discrete access point model. The current state of the art model typically assumes that each antenna typically feeds its own separate RF transmit / receive chain, which is then processed using a linear decoder as appropriate for MIMO transmission. When multiple switched antennas can be selected for an RF chain, they are aggregated into an antenna array, co-located at the access point, typically spaced or angled to take advantage of a particular directivity or intentional interference pattern created by a fixed radiator. In such cases, the antenna positions and angles are assumed to be strictly fixed and constrained to a narrow range of collocation. For example, an array based on beam directionality will only work if enough overlapping directional antennas can be placed on a fixed assembly such that overall coverage can be maintained. This assumption permeates down to the process and software used to make such selections. In the flexible wireless networking system described herein, antenna selection is not co-located in a fixed array. The selection in FIG. 22 is generally not a beamforming selection. For example, individual antennas 2106 are potentially many yards apart, with distances and orientations not necessarily known in advance and infinitely variable based on the geometry of the installation (folds to move around pipes, turns around wall corners, etc.) The choices the system makes are to position, move, and shape cells over broad regions of space, and thus are not limited to beamforming or similar co-location operations.

[0059] Note that the choice made in FIG. 22 may ultimately not be about any particular client, but about the projection of the antenna pattern over a significant region of space and what the results will be. The goal of a strip deployment is usually to cheaply and easily fill an area, e.g., a building, with an infinite number of radiation options. There will be multiple radios, and there will be multiple collision regions as multiple radios compete for upstream and downstream access. If two radios are on the same channel, the downstream of that radio and the upstream transmission of the client can interfere with each other. The interference always occurs at the receiver, depending on the ability of the receiving radio to reject unrelated competing simultaneous transmissions as noise. This in turn depends on a variety of factors, such as the transmit power of the signal, the coding used, the nature of the training preamble and headers employed in the protocol, etc. If the radios are on different channels but adjacent, there is also the issue of spillover, since in many cases the transmitter actually spills power outside its designated channel band, with a well-understood rollover function that can cause interference between clients that are adjacent to each other on those different channels for range or load balancing purposes.

[0060] In some cases, multiple radios may each have access to multiple antenna options. To simplify the analysis, one can imagine an antenna strip made of very simple antenna patterns that repeat for each radio. An example of these antenna patches is illustrated in FIG. 23A-B. Specifically, FIG. 23A shows a mixture of radiation shapes and directivities that produce the radiation pattern 2310 shown in FIG. 23B, which has a flower shape. Each element 2305 of the pattern of antenna patches shown in FIG. 23A produces one of the overlapping shapes of the overall flower pattern 2310 shown in FIG. 23B, and although switches and lines are not shown for clarity, each patch can be added or removed independently from the overall pattern. The patches are not necessarily in a fixed arrangement, and patch elements may be at distances greater than the collocation distance from each other within a patch. Furthermore, one can imagine the strip 2300 carrying that repeating pattern of radios with patches on the ceiling, both in the top and bottom halves of the horizontal rectangular space (e.g., a building) to be covered.

[0061] FIG. 24 illustrates an example of a flexible wireless networking deployment, resulting in the illustrated pattern 2400, which allows the deployment for coverage of an entire building (shown with respect to floor plan 2410) due to channel reuse. In this example, each channel is shown with a solid line for channel A (e.g., as in pattern 2402), a dashed line for channel B (e.g., as in pattern 2404), and a dotted line for channel C (e.g., as in pattern 2406). Unfortunately, even in this illustration, which is sparser than may be the case in real life, the overlap within the overall pattern 2400 is significant when all are turned on. Strings with larger antenna patterns (such as those with multiple remote antenna locations) or strings with a higher density of radios will cause even greater overlap. In some deployments, this overlap is a clear advantage. However, in many multi-channel deployments, it may be advantageous to avoid such large overlaps.

[0062] FIG. 25 shows how the overlap in the overall pattern 2500 can be reduced. For example, a portion 2502 of the pattern 2402 for channel A, a portion 2504 of the pattern 2404 for channel B, and a portion 2506 of the pattern 2406 for channel C are kept, while other portions of these patterns are removed. (Note that in FIG. 25, the background floor plan 2410 has been removed for clarity.) There are advantages to doing this. For example, by not spreading the coverage pattern too much, both co-channel and adjacent channel downstream interference is minimized. Furthermore, because the cells are sharp, beacons and other downstream traffic from the cells are prevented from spreading out too far in space, preventing clients from remaining attached to cells that are farther away. This makes the clients much more likely to change channels, thereby preventing them from causing co-channel interference with a third cell. As a result, the coverage becomes more predictable. Additionally, if there is a concern about signal leakage for other reasons, such as accidentally picking up more noise from a machine such as a microwave that would otherwise be in the coverage pattern, the problematic patch can be turned off.

[0063] Note that the coverage pattern itself is arbitrary. For example, Figures 26A and 26B show a similar analysis using a simpler, non-realistic, overlapping circle pattern. In this example, a portion 2602a of the pattern 2602 for channel A shown in Figure 26A, a portion 2604b of the pattern 2604 for channel B shown in Figure 26A, and a portion 2606c of the pattern 2606 for channel C shown in Figure 26A are retained in Figure 26B, but other portions of these patterns are removed from the overall pattern 2600, etc.

[0064] As mentioned above, even the example patterns shown in Figures 24 and 25 are simplistic compared to reality. The extreme density and pattern options made available by the flexible wireless networking system described above can create many times more opportunities for detection, resolution, and optimization in the densest form than systems implementing the traditional discrete access point model. In traditional deployments, the system lacks the flexibility to make meaningful changes to the deployment. Due to the cost and complexity of installing individual base stations, the number of base stations is severely limited, and so they are placed under pressure to be as far apart as possible to reduce wasted resources, and as close together as possible to ensure smooth coverage. Most base stations have only two configuration options: a power level decision to avoid exacerbating interference by extending coverage too far, and whether to completely turn off a completely useless base station. The software algorithms used to make these decisions in traditional deployments are very limited due to the lack of useful information available and the flexibility to do something about it. However, with the degrees of freedom offered by the flexible wireless networking systems described herein and the multiple antenna patterns they contain, disclosed herein are methods for using those degrees of freedom to precisely allocate antennas and radio resources to maximize coverage.

[0065] 27 illustrates one embodiment of a system and method for software optimization of a flexible wireless networking system. In the illustrated embodiment, the system includes one or more radios 2714 (two are shown) and one or more switches 2716 that control one or more antennas 2718 transmitting and receiving into a space with potentially multiple wireless client devices 2720. The different antennas 2718 between the radios 2714 may or may not be within range to communicate with each other and with various ones of the wireless client devices 2720.

[0066] In typical operation, multiple radios 2714 may operate on the same channel or different channels, up to the number of channels available before forced reuse. However, radios 2714 may change to the same channel to ensure communication. Furthermore, in protocols such as Wi-Fi, the wireless client device 2720 itself periodically sends requests (such as probe requests) on channels different from its operating channel to ensure that as it moves through the environment, it understands what services are available on different channels at the current moment. In some embodiments, one goal is to sharpen cell boundaries and reduce the number of service options that a wireless client device is exposed to at any given location in space to one strong option, which may result in better, more predictable performance.

[0067] One approach to gather information to perform any optimization or configuration is to maintain a list or table of neighbors for each antenna, shown in FIG. 27 as antenna neighbor table(s) 2702. These neighbor tables differ from the prior art in at least two important ways. First, in a discrete access point (pre) environment, the neighbor tables are often learned per access point itself, rather than per antenna. MIMO environments require simultaneous antenna transmissions in any case, and typical neighbor exchanges (such as from 802.11k) are largely agnostic to the particular antenna selection, which is rather considered as a PHY layer function under the purpose of MAC layer knowledge building. The exchange of reports does not include a way to identify or isolate which antenna did what under a given MAC address. Second, approaches to neighbor detection in conventional models were designed well short of the explosion of antenna degrees of freedom possible in the flexible wireless networking system described herein. As mentioned above, conventional deployments at best use antenna arrays that are fixed collocated assemblies and are not flexible over the spatial domain (both in configuration and in the literal bendability of some embodiments).

[0068] In some embodiments, the antenna proximity table 2702 contains information indicating if and how two antennas share a radio environment, including direct and indirect reach of each other, and what the sharing effects may be. There are many ways to arrange this.

[0069] Some embodiments populate the neighborhood table using telemetry analysis, shown in FIG. 27 as telemetry analysis 2710. Passive telemetry is information obtained from the typical operation of the network without making an active effort to collect specific information. One type of passive telemetry stream is that of an active antenna receiving information known to be transmitted by other antennas. Specifically, the system maintains a list of all antennas that are selected to transmit, shown in FIG. 27 as antenna transmitter selection list 2704, and what their unique identification is. In some embodiments, this unique information may be a transmitter address, such as an Ethernet MAC address. In other embodiments, it may be a PHY level indicator, such as a BSS coloring identifier, which when combined with a spatial color ID map (maintained or accessed by the system), may be used by one receiver to determine which particular transmitter radio was used when a particular color ID was received. In some embodiments, the unique information may also include a spatial stream identifier. In some embodiments, the information recorded may be a transmission from a system radio that was received by another system radio. In other embodiments, the recorded information may be traffic directed to the system radio by a third device detected by the second system radio, or traffic directed to an entirely different radio by a device that is also known to be in communication with the system radio. The system may also maintain a list of all antennas selected for reception, along with their unique identities, shown in Figure 27 as antenna receiver selection list 2706. Associated clients table 2708 is described below with reference to Figures 29A, 28B, 29C, and 37.

[0070] To detect overlap between systems, some embodiments may use an antenna proximity table 2702 to match promiscuous reception of transmissions from individual receivers to a transmitter and record specific signal information such as signal strength, SNR, data rate, number of spatial streams, or even channel matrix or summary data therefrom. For Wi-Fi, one embodiment collects telemetry streams from every radio for every frame transmitted by any other radio in the system, as determined by the transmitting MAC address. Telemetry analysis may be performed on this data. In some embodiments, some or all of the data representing the receiver configuration and transmitter configuration is recorded for each detected transmission. This is important because both the transmitter configuration and the receiver configuration may change in real time. In a MIMO system, the transmitter can adjust the weights it assigns to each transmit chain, and the receiver constantly adjusts the receive weights to optimize the transmit SNR of each stream. In a non-beamforming environment, the transmit weights are typically static (avoiding per-packet power control) and the receive weights are merely optimized, thus revealing information about the changing channel. However, in a beamforming environment, transmitters may be steered away from the system receiver towards the intended client, in which case a lower received signal strength may not be a good indicator of reduced overlap. Therefore, beamforming should be used with care. System-side beamforming usage may be determined by querying the radio. For non-system transmitters, it may be possible to infer beamforming usage by observing beamforming management traffic such as channel sounding.

[0071] Because by design multiple antennas can be used to transmit and receive, the received information may apply to antenna combinations rather than to individual antennas. FIG. 28 shows an example of this effect. In the example shown, transmissions come from multiple antenna combinations 2804, each of which includes multiple antennas 2802, and are received by multiple antenna combinations 2806, each of which includes multiple antennas 2808. These combinations may potentially overlap. For example, they will overlap in embodiments that dynamically adjust the combinations during an analysis event. In some embodiments, this may be avoided by enforcing a static mapping that prevents changes during an analysis event and has a non-overlapping set of antenna combinations. However, in those cases, there may not be enough information to understand the contribution of the individual antennas to the combination.

[0072] With overlapping antenna sets, one possible challenge is determining the potential contribution of each antenna to the commonality. However, because there is commonality, it is possible to determine mutual information using common reception of the transmission, or separate reception of similar transmissions. In some embodiments, it is possible to share one antenna between two radios, using a multi-lane strip to connect two radio chains to that antenna. In some embodiments, doing so may combine all of those antennas into one shared antenna configuration, and the actual difference to the channel response of that set is based on which radio is accessing it. For embodiments that do not allow such multi-radio connections (either by hardware or software), overlapping antenna sets may instead be generated by the system deciding to switch set assignments during the analysis period.

[0073] 29A and 29B are diagrams illustrating two similar transmissions picked up sequentially, with the same transmitter, same channel weighting, same location, same coding and data rate, and same power. Specifically, the first transmission is picked up by one configuration of antennas 2900 with two overlapping antenna sets 2902 and 2904, as shown in FIG. 29A, and the second transmission is picked up by a different configuration of those same antennas 2900 with sets 2906 and 2908 with different overlap, as shown in FIG. 29B. In this example, signal strength may be used as a determinant of how well an antenna "hears" a client. SNR, channel weighting, loss rate, error rate, or other factors may be used as determinants in other embodiments. FIG. 29A shows antennas 1-5 of set 2902 and antennas 4-8 of set 2904, and FIG. 29B shows antennas 1-6 of set 2906 and antennas 5-8 of set 2908. In this example, the reception in the first case has signal strength S1, and the reception in the second case has signal strength S2. Since the two transmissions are similar, if S1 and S2 differ, one may assume that the difference is due to the losses of antenna 4 and the gain of antenna 6 on the one hand, and the losses of antenna 6 and the gain of antenna 4 on the other hand. This gives evidence of the values ​​of antenna 4 and antenna 6.

[0074] Assuming that we use only antennas 1-6, and only the first (leftmost) set in each case, the only difference is that antenna 6 was previously unused (in set 2902) and will be used later (in set 2906). If S2>S1, then we give the information that, for that particular client location, antenna 6 provides a gain of S2-S1 in the antenna combination. Here, the data is clear. We can think of the coverage pattern of antenna 6 as overlapping better with the transmitter. This particular information corresponds to a proposal P({1...5}+{6}→S1+(S2-S1)|T), i.e., that adding antenna 6 to set {1...5} will increase the gain of the antenna set for transmitter T by S2-S1. Note that this is a computationally representable proposal. In the example shown in FIGS. 29A and 29B, proposals P({1...5}+{6}→S1,1+(S1,2-S1,1)|T) and P({4...8}-{4}→S2,1-(S2,2-S2,1)|T) are generated, where S1,1 and S1,2 are a first set of signal strengths for the first and second transmissions, respectively. Such proposals may be mathematically solvable or approximately solvable in the presence of conflicting information. Some embodiments may use these types of representations to generate satisfiability problems. Other embodiments may use these representations to generate optimization problems with accompanying utility metrics. Still other embodiments may use different proposal structures that are tailored to the particular problem or problems being inferred, such as whether a configuration provides adequate coverage or whether a configuration uses resources appropriately.

[0075] Signal strength is not the only telemetry accessible, nor is it the only proposal that can be constructed, and minimizing overlap may not be the only desired optimization. In some embodiments, signal strength telemetry is used to provide different overlap balances. For example, some embodiments maximize coverage overlap, while others maximize overlap subject to limiting constraints (such as available power or network resources). Some embodiments collect data rate telemetry. Some do this with the goal of optimizing around data rate availability, thereby maximizing the data rate available for the deployment. For example, data rate may be related to signal strength, but also to channel width availability and spatial dimension availability (both characteristics of the environment and interferers therein). The optimization may be joint, or approximate, or balanced in various embodiments. Since more telemetry is better in most deployments, the neighborhood table is operated as a data lake, and its contents are available for any analysis. In some embodiments, the neighborhood table is subject to statistical analysis. Longer time period analysis may be performed. Analysis of these different time periods may be used to guide the configuration.

[0076] There are multiple ways to correlate client transmissions between different receiving system radios. One such way is to use tightly time-correlated events. An example embodiment uses probe requests from time-bound clients. Clients that perform channel scanning typically do so in a tight loop so as not to disrupt their own service, where they scan as many channels as possible to discover what is available while transmitting probe request frames. Such probe request frames are typically transmitted at the same power level, subject to regulatory channel limitations. Thus, an example embodiment may perform reception using different probe requests from the same scan (which are identical when they are on the same channel, but close in time when they are not).

[0077] Another source of information for correlating client transmissions between different receiving system radios are 802.11k neighborhood reports, which can be used in combination with system antenna status. For example, when the network requests a neighborhood report from a client, it requests the client's scan database provided by cooperating clients. This provides information about what the client can hear, which is the opposite of what would be obtained if the network were to simultaneously listen to the client's transmissions. Here, instead, the client tells the network about all the other transmitting system radios it has heard. However, for simple transmissions, this reversed information is likely to be similar. For example, beacons are typically transmitted in a simple spatial stream configuration, and signal strength is unlikely to vary with location in the environment, and therefore signal strength is more likely to correlate with pattern response than environmental multipath. Thus, some embodiments use the neighborhood response to populate the antenna neighborhood table. This can be useful when the neighborhood reports are fresh or time-bound enough to reveal quality differences from the client's perspective when antenna sets change and the network undergoes changes, or when a client moves to a location with a different antenna radiation pattern.

[0078] Some embodiments use active telemetry to correlate client transmissions among different receiving system radios. Active telemetry involves consciously coordinating the network's operation to collect this information, while recognizing that this type of testing may in some cases result in minor or major service interruptions. One such method is for some radios to briefly hop to the same channel and simply snoop on the reception. This can be easily done in over-deployed networks where many radios are already collecting promiscuous telemetry. In those cases, the effort may involve a continuous change of surrounding promiscuous radios to listen and correlate simultaneous receptions as described above, using identical frames if desired.

[0079] Another method for correlating client transmissions between different receiving system radios involves actively provoking intermediate stations such as clients. Clients associated with a BSS (e.g., in the associated client table) can obviously be pinged by that BSS. But nearby radios can also ping clients. Depending on the protocol, clients usually respond to transmissions from any device at the lowest level of the protocol stack, even if it means ignoring transmissions at higher levels. For example, in Wi-Fi, a device that receives an RTS ("request to send") addressed to it responds with a CTS ("clear to send"). The RTS does not have a BSSID. Also, even if it ignores RTSs that are not from a base station, it can forge the source address of the RTS and use a unique duration field length that is not real. Since the responding device subtracts the time already taken and sends it in the CTS, it is also possible to calculate the RTS duration to send so that the CTS comes back with a meaningless number such as "1" that is too short for an actual transmission, for example. In some embodiments, a radio that wishes to determine the geometry of its antenna may send pings from other clients in such a manner. In one embodiment, which may be useful for serving radios that do not maintain promiscuous mode while operating as a base station, for those that filter traffic, the injecting radio may choose to forge frames on behalf of the filtered base station radio so that the client's response meets the filter criteria. For example, if the base station radio is a filtered Wi-Fi that only takes management and data frames, the injecting radio could inject a mandatory response frame (such as an action frame) and the receiving radio would be able to get that response and issue it for correlation.

[0080] Whether the client is forced to transmit or the network is chatty enough to opportunistically use its own transmissions, in some embodiments the test radio may select which antenna (and set of antennas, if that information is desired) to target, perhaps rapidly scanning through antennas and combinations of interest, and sending pings to various clients. In some embodiments, an active measurement system, such as the active measurement 2712 shown in FIG. 27, may be arranged to coordinate multiple radios and thus make specific comparisons between antennas. For example, the active measurement system may issue a command to the effect of "Radio 1, try your antenna 1. Radio 2, try your antenna 2. Tell us what you hear so we can decide which is better." Some embodiments may use a simple search. Other embodiments may use an exponential search limited to closer or nearest neighbors. Still other embodiments may search already collected information to determine holes or inconsistencies (e.g., rank the suggestions recorded above in order of inconsistency from the solution and retest the most inconsistent). This search may occur in the background while the network is operational, and different methods may be mixed. In busy networks, telemetry may be sufficient without pinging, but pinging may always be available.

[0081] Some embodiments infer the location of the client while performing this process. Some embodiments do this to augment the data with additional quality information. For example, data from a client whose location may be fixed can then be compared to data from physically nearby clients because the physical location of the client tends to correlate with the channel response in most environments, even though the channel response itself changes rapidly as people move around. Some embodiments can use a more location-friendly antenna pattern to first derive the location of the device before moving on to testing more exotic patterns. A more omnidirectional antenna pattern may allow some location detection algorithms to do better triangulation. In some embodiments, the location positioning may be switched to another protocol, such as Bluetooth, to perform the location determination. For example, if the positioning radio is in a known location relative to the serving radio (e.g., if the Bluetooth radio is on a separate antenna co-located with the Wi-Fi antenna), the derived location may be of high quality for direct use.

[0082] Overall, some embodiments piece together the acquired information to generate a learning phase. The learning phase may be applied to the entire network, or piecemeal by splitting the network and working on one part at a time. In some embodiments, the learning phase involves taking the area in question, setting various radios to the channel in question (based on which channel will test the most clients, or will disrupt the serving radio the least, or as part of a scan sequence, etc.), and then rapidly running tests to identify the neighborhood. Some embodiments may perform these tests using only non-serving radios, while other embodiments may hand off clients onto another radio to bring them up for service, to allow testing of the non-serving radios. Some embodiments may instead issue commands (such as by issuing a long RTS frame or announcing a quiet period) that take up the airtime needed to perform the search.

[0083] Some embodiments may use "winking" explicitly, where the term "winking" refers to the operation of a radio with multiple antennas in a radio chain experimentally adding or subtracting from its set and collecting its data, e.g., adding or subtracting one antenna at a time to determine the impact of each such change. If a radio in use is in service and does this winking while transmitting, it may continue to send real traffic to clients or may send dummy traffic (such as QoS NULL frames or other messages) while performing this test. Similarly, while receiving, the radio may perform winking by first eliciting a futile response (e.g., by sending action frames for information it already has or that is of low value) and then ensuring that it winks at the appropriate time. In some embodiments, winking occurs by making small adjustments (e.g., adding or subtracting one antenna) every now and then to see if it makes a difference. In some embodiments, the winking becomes more aggressive and actively searches for different systems. Winking does not need to induce client traffic and can be optimistic. In some embodiments, winking is performed probabilistically using an adjustable probability threshold to introduce a small but meaningful long-term non-destructive (or bounded mildly destructive) comparison measurement opportunity.

[0084] Strict timing coordination between the radio and the telemetry system may not be necessary. In some embodiments, the antenna switch is changed somewhat slowly depending on the radio's air activity. Nevertheless, the data can be tightly correlated in the analysis because in almost all cases the radio generates a very good time counter for the receive or transmit record, which can then be correlated with a tight counter (such as a CPU cycle counter on a GPIO CPU) that controls the switch.

[0085] In some embodiments, the neighborhood may be limited to system resources. In some embodiments, clients of the system may also be maintained in a neighborhood table, thus establishing a more global topology. In some embodiments, this may be extended to non-associated devices that may also be monitored.

[0086] Neighborhood maintenance may vary in different embodiments based on the goal. As mentioned above, the received information may represent suggestions regarding the topology of the network. The suggestions themselves may be binary or weighted. For example, changes in signal strength may be used to generate the weights. There are several known algorithms for generating (usually approximate) sufficiencies of logical suggestions. However, in some embodiments, the neighborhood suggestions are taken in a Bayesian manner where each suggestion adds some information about the ultimate problem, such as whether antennas 1 and 2 overlap to some measure. Since that information is often obtained indirectly, such as through intermediate stations such as a particular transmitter T as mentioned above, these embodiments use each piece of information to generate a Bayesian network that can be queried, the methods of which are known in the art. Some embodiments where exact Bayesian inference is not required may use simpler machine learning (ML) models to generate oracles to understand the relationships and query potential changes.

[0087] FIG. 30 illustrates a method 3000 for dynamically configuring a flexible wireless networking system using machine learning, according to some embodiments. In certain embodiments, one or more of the operations illustrated in FIG. 30 may be performed by a resource manager, such as the resource manager 1940 illustrated in FIG. 19. In the illustrated embodiment, the telemetry process 3002 generates weighted suggestions 3004, which are added as inputs to the ML model 3006. In some embodiments, the ML model 3006 observes each new suggestion presented along with a measure of the network quality of the area affected by the suggestion, learning, for example, using backpropagation, that a particular suggestion is performing poorly. After sufficient training, the ML model 3006 can present hypothetical suggestions, such as in the form of partial or complete configurations, and generate a predicted quality of outcome for each configuration, shown in FIG. 29 as predicted configuration result 3008. The configuration manager 3014 may then direct the decision between different proposals based at least on the predicted configuration results 3008, such as by outputting a configuration decision 3012 and / or adding one or more configuration proposals 3010 as inputs to the ML model 3006. In some embodiments, the proposal itself is made by keeping a record of the worst performing parts of the network, shown as a faulty telemetry record 3016, and the weighted proposals 3004 associated therewith. In some embodiments, the faulty telemetry record 3016 may also include the predicted configuration results 3008 for the proposals for which the predicted results are faulty. The configuration manager 3014 may use this record 3016 to identify parts of the configuration that need change, and in some embodiments may constrain the search space of proposals to areas most in need of help, or to the strongest antennas impacting quality. In some embodiments, machine learning may be employed directly on the identity of the antennas, with the quality impact backpropagated as a prediction, to derive a learned response of the significance and / or importance of the antennas for different configurations. The configuration manager 3014 can then directly interrogate each antenna to learn its importance and use that information to configure a satisfactory set of antennas.For example, the inclusion or exclusion of a particular antenna in the proposed configuration may affect the predicted performance of the proposed configuration, positively or negatively, in negligible or significant amounts. In some embodiments, a machine learning model may be trained on the configuration set itself as input, and may maintain a longer term record of what worked and what did not. In this type of architecture, the ML model that maps the configurations or parts of them to qualities may require exploration to discover which configurations are reasonable or preferred. In some embodiments, a generative adversarial network is used to generate configuration possibilities. In some embodiments, an ML model 3006 that may be trained on the proposals is used to generate a discriminator (as shown in FIG. 37) and a generator is combined. In various embodiments, the generator may generate results from noise input, from past configurations, or from the proposed configurations.

[0088] Some embodiments can use the antenna proximity table to automatically prune or shape the resulting coverage area. This automatic pruning can perform the same function as manual pruning, but without the need for human intervention. There are several different dimensions that can operate simultaneously to affect pruning. The first is the location of the cell. For example, as mentioned, it may be desirable in some configurations to keep the cells compact. This may be true for several reasons, such as to provide predictable coverage and to avoid interference. However, the flexibility of the wireless networking systems described herein may allow the system to assemble antenna sets from patches at various possible distances from each other. This may have both advantages and disadvantages with respect to cell assembly. One advantage of increasing the number of antennas switched together on the same antenna set is that more antennas may provide different gains, impedances, and directivities. More antennas can increase coverage by combining patterns. For example, adding far-away antennas can create multiple opportunities for reception at those distances. Furthermore, in some embodiments, the pattern may be arranged with parasitic elements so that very high gain antennas, operating in principle similar to Yagi antennas, can be assembled on the fly using a switch to select those antennas near the parasitic or reflecting elements. In some embodiments, non-radio elements may be switched on and off with ground to affect their reflectivity. On the other hand, all antennas may have different channel responses. Antennas within a precise distance of each other (typically short distances, since precision is difficult with length) could potentially inadvertently form an antenna array, and thus the combination of two antennas could result in an antenna with a different directivity and possibly unpleasant fading pattern than either of its parents. Therefore, in some embodiments, antennas may be arranged locally to avoid this situation.In some embodiments, each antenna is given a higher directivity, thus ensuring that an area is less likely to be equally within range of both antennas and inadvertent interference fading can be avoided. In some embodiments, the interference pattern is known and additional antenna segments can be created to cancel the interference. For example, if two antenna patches interfere horizontally to create an interference pattern, one additional antenna may be placed in phase between the two antennas to create an interference pattern that is 180 degrees out of phase using an excessive number of segments. In some embodiments, both interference and non-interference patterns exist and a choice can be made between the two. In some embodiments, the interference potential of different antennas may be added to the suggestion logic of the neighborhood table. For example, in some embodiments they may be absolute assertions (e.g., skip the combination or drag along an additional antenna to cancel the interference). In other embodiments, the interference may be recorded and taken into account in the weighting.

[0089] Beyond interference, when there is distance between antennas, they will have different first order signal delays. In some embodiments, the signal delay from any set of antennas is required to be within some tolerance, often based on the ability of the receiving radio to reshape the delay pattern (number of taps, etc.) into an aggregated overall high power. In some embodiments, the antennas on the strip are a priori divided into specific repeating patterns from which only antennas in the repeating regions may be combined into one or more sets. An example of regionalization is illustrated in FIG. 31. In the illustrated example, the aggregated local antenna regions allow the configuration manager to take local characteristics into account. While some embodiments may individually aggregate each of the solid line antenna patterns into one or more antennas, the embodiment illustrated in FIG. 31 specifically limits the options to any one of the dashed circled sets 3102 defined for the collection of antennas 3100. Thus, if any antenna in the first set (3102a) is used in the set, only elements from set 3102a may be used, the rest are prohibited. This constraint may be enforced by software or hardware restrictions. An advantage of doing so may be that the interference and directional characteristics of a circled antenna set 3102 may be well understood during manufacturing or design, but the characteristics between segments in different circled antenna sets 3102 may not be well understood. For example, installation instructions may require that the strips not fold in the middle of the circled regions (or even have a more rigid backing to prevent folding there), yet still allow the strips to fold more freely between the circled regions. In some embodiments, the strictness of locality may not be explicitly enforced, but may be introduced as a weight or priority that can be overridden depending on optimization or sufficiency utility metrics.

[0090] In some embodiments, cells are kept small so that they can be relocated some length down the strip in one move. For example, FIG. 32 shows a strip 3200 with repeating patterns 3204a-3204n that span the length of the strip 3200, including between radios 3202a and 3202b. In these embodiments, the system provides a great degree of flexibility in where radio resources are placed on the radios without requiring each potential location to have a radio to select a local set to place in service. FIGS. 33A, 33B, and 33C show several different possibilities for selection of distance for the strip 3200. For example, FIG. 33A illustrates a scenario in which the third of the repeating patterns (shown as 3204c) is selected for radio 3202a and the fourteenth of the repeating patterns (shown as 3204n) is selected for radio 3202b. Figure 33B illustrates a scenario in which the first repeat pattern (shown as 3204a) is selected for radio 3202a and the eighth repeat pattern (shown as 3204h) is selected for radio 3202b. Figure 33C illustrates a scenario in which the first repeat pattern (shown as 3204a) is selected for radio 3202a and the fourteenth repeat pattern (shown as 3204n) is selected for radio 3202b.

[0091] The selection of this particular pattern may be similar to the combination of the examples shown in FIG. 24 and FIG. 26A, B, where each circle in FIG. 26A, B represents a flower in FIG. 24. This provides the possibility of a joint optimization, the goal of which is to find the best locally constrained "flower" set of radiators to use while tapering that set to perform the optimization. Some embodiments may perform the optimization in two steps, first selecting the best cohesive set of radiators and then tapering that set, rather than as a joint optimization. Some embodiments may perform a backtrack to reselect a different first set of radiators if tapering becomes too infeasible. Some embodiments may perform a proper joint optimization, where a local cohesive constraint is added to the satisfiability or optimization engine.

[0092] In some cases, changes in an operational network may be advantageous. For example, it may be advantageous to perform seamless roaming in a network simply using physical layer resources. Seamless roaming refers to the goal of causing minimal client disruption as wireless client devices move in the network. In a conventional deployment, the antenna radiation pattern of a base station may not move. Thus, as a wireless client device moves away from a base station, the client device experiences reduced availability of the base station's radio resources, leading to a point where the client feels it must seek an alternative (or is instructed to do so by the network). At that point, the client performs a handoff procedure to a new base station. Both the client's search function (e.g., scanning) and the handoff mechanism take valuable time and require management traffic overhead. In general wireless networking, many measures have been taken to limit these effects, but they still exist. Furthermore, in networks with large client groups, clients may make decisions based on factors other than signal strength, such as the responsiveness of an overloaded base station and available airtime. This can lead to unpredictable emergent behavior in the client pool, such as inadvertent swarming, where clients make similar simultaneous decisions to flee to lightly used base stations, resulting in flapping or thrashing as the base station becomes overloaded, or paradoxically self-reinforcing bursts of management traffic.

[0093] However, the flexible wireless networking system described herein may provide physical resources to seamlessly move radio resources with wireless client devices. Figures 34A, 34B, and 34C are exemplary embodiments including individual switchable antennas 3408 attached to the same radio 3402 as a client device 3404 moves relative to and along the strip 3400. In this example, only one radio 3402 is turned on at a time. As the client device 3404 moves along the strip, the switches of the various switch / controller pairs 3406 are changed to ensure that the antenna 3408 closest to the client device 3404 is the only antenna serving the client device 3404. For example, Figure 34A shows a time when the client device 3404 is served by antenna 3408a, Figure 34B shows a later time when the client device 3404 is served by antenna 3408b, and Figure 34c shows an even later time when the client device 3404 is served by antenna 3408c. In this particular case, the client device 3404 will not experience a handoff along that portion of the strip 3400. Instead, the client device 3404 will experience the network following all the way. As one moves further away from the radio 3402, there may be a small drop in signal strength, in which case a power amplifier can be employed to boost the signal, if desired.

[0094] In some embodiments and under certain circumstances, multiple antennas may be switched on at once. In fact, in some embodiments, all of the antennas may be switched on at once. Doing so may create a "leaky cable" effect, which may increase the antenna delay spread or alter the MIMO channel characteristics, potentially overloading a rake receiver or similar signal processor in the radio. However, it may be reasonable to limit the antennas to some constrained delay loss pattern, or to a directional or spatially diverse set of antennas.

[0095] 35A, 35B, and 35C illustrate a handoff scenario using a limited set of radiators, according to some embodiments. Again, to avoid clutter, the switch fabric has been removed and only the radiators are shown. This configuration combines the handoff problem with the constraint of using a coherent set of radiators to avoid delay spread or ensure coverage. In this example, FIG. 35A illustrates a time when a client device 3504 is served by an antenna configured with a set of antenna elements on a strip 3500 shown as 3506a, FIG. 35B illustrates a later time when the client device 3504 is served by an antenna configured with a set of antenna elements on a strip 3500 shown as 3506f, and FIG. 35C illustrates an even later time when the client device 3504 is served by an antenna configured with a set of antenna elements on a strip 3500 shown as 3506n.

[0096] In some cases, the optimization or sufficiency solutions described above may be used to handle changes in client location. In embodiments that perform real-time analysis, this works well. Some embodiments go further and trigger reconfiguration in response to client indications of motion or impending handoff. In various embodiments, there may be various ways to determine when and if a wireless client device should handoff from one or more currently active antennas, where signals are transmitted or received on behalf of the wireless client device in a first coverage area, to at least one alternative antenna that may have a second coverage area. For example, in a system with much more antenna resources spatially than there are radios, there may not be enough passive radios to find better coverage for an individual client. In such cases, some embodiments may use location tracking to determine when a client may be approaching a cell boundary and adjust its client table in response. Other approaches may include looking for probe requests on neighboring radios that come in with higher signal strength than the serving radio or higher signal strength than expected, looking for action requests for handoff advice, or looking for a drop in data rate or signal quality as an indication that the client is moving towards the end of its coverage and needs reconfiguration. In many of these embodiments, the system deploys the information quickly enough to meet the needs of the mobile client. However, some embodiments may not perform such optimization in real time or may not be tailored to deploy such information in real time to enable such fast reconfiguration. For those embodiments, a multi-tiered approach with a faster handoff manager function integrated with slower optimization may be used to address this issue.

[0097] FIG. 36 illustrates one embodiment of a system and method for software optimization of a flexible wireless networking system including a handoff manager 3602. In some embodiments, the handoff manager 3602 may be similar to the handoff manager 1920 illustrated in FIG. 19 and described above. Other elements shown in FIG. 36 may be similar to those illustrated in FIG. 27 and described above. In the illustrated embodiment, the handoff manager 3602 is used to make short-term antenna configuration changes to handle wireless client device movement while utilizing an antenna proximity table 2702. The handoff manager 3602 checks the associated client table 2708 and uses that information to query or filter telemetry data to ensure that it knows when a client may need a handoff adjustment. For example, telemetry data indicating an increase in a client's error rate, a decrease in signal strength on an associated radio, or an increase in signal strength or polling by clients on other radios may all be valuable in the reanalysis. In some embodiments, the handoff manager 3602 also performs short-term active telemetry analysis (shown at 2710) using an active measurement system 2712 to initiate changes or scans to the antenna switch 2716 to quickly determine availability of improved configurations of resources, as described above. In some embodiments, the active measurement system 2712 may be shared between longer-term neighborhood discovery and shorter-term handoff discovery. Note that neighborhood discovery may be primarily focused on developing spatial awareness of the interactions of antenna patterns between antennas and can use client information to do so, while handoff discovery may be primarily focused on finding optimal client associations to each antenna. In some embodiments, neighborhood discovery and handoff discovery may share similar operations and data, and indeed the handoff data set may be embedded within the neighborhood table.In other embodiments, neighbor discovery and handoff discovery may share many commonalities, but based on tuning parameters and installation or configuration choices, handoff manager 3602 may need to fill in some gaps to achieve short time horizons and minimum responsiveness requirements where the entire neighbor discovery process cannot be adjusted.

[0098] FIG. 37 illustrates one embodiment of a method 3700 in which a handoff manager 3702 is integrated with a configuration manager 3014. In some embodiments, the handoff manager 3702 may be similar to the handoff manager 1920 illustrated in FIG. 19 and described above. Other elements illustrated in FIG. 37 are similar to those illustrated in FIG. 30 and described above. In certain embodiments, one or more of the operations illustrated in FIG. 37 may be performed by a resource manager, such as the resource manager 1940 illustrated in FIG. 19. The illustrated embodiment includes a composite configuration generator and identifier 3704. However, a more general arrangement as disclosed above may be used instead, in which the handoff manager 3702 has access to the configuration manager 3014 and the telemetry process 3002. The handoff manager 3702 processes the telemetry from 3002 (including possible active measurements) as is conventional to determine the active needs of the mobile client device. The handoff manager 3702 adjusts the weighted proposal 3004 when a handoff seems warranted, etc., to indicate the movement of the client device or an inference of its motion over the spatial radio environment. For example, a location tracking system determining that a client device is within coverage but has impaired signal quality everywhere can update the proposal associated with that client device, or introduce a new proposal suggesting that another antenna tested with a transmitter currently in a similar location to the client device may perform better. The handoff manager 3702 may then request the configuration manager 3014 to generate a new configuration decision 3012 and / or one or more configuration proposals 3010. In some embodiments, the configuration change region may be constrained to prevent changes at a distance (butterfly effect). Such constraints may include attenuation or penalty weights for changes at a distance, constrained proposals (which may be temporary and apply only to the handoff requested reconfiguration, but allow global reconfiguration when necessary), or constraints that limit the configuration to a smaller region and merge the smaller region back into the entire configuration set.

[0099] In some embodiments, the handoff manager may not force a new computation of the reconfiguration, but may select a new configuration from a pre-configuration set to determine a more optimal mapping. In these embodiments, the configuration generator may not be activated for handoff or reconfiguration. Instead, the discriminator may be presented with several options that take into account the moving client or the client's service needs, and may select a new configuration from among those options. Some embodiments may provide limitations or constraints on the reconfiguration to prevent rippling in distance. In some embodiments, the pre-computed options may be at the coherent local antenna set level. In such an embodiment, the handoff manager may drive the system to select between coherent sets, but may leave the selection of each antenna in the set to a wider neighborhood optimization. The effect of all these options may be to allow fast selection between already constructed feasible options that properly consider interference and coverage, or whatever the optimization criteria, and then independently ensure better handoff behavior. A slower optimization process may continually update the more feasible sets. In some cases, the more feasible set represents a top level of nearly interchangeable options from which a second optimization can occur to maximize utility as the mobile client moves. In some embodiments, further constraints may be provided to attempt to maintain the MIMO channel configuration. For example, in the scenarios shown in Figures 35A, 35B, and 35C, the individual sets of antenna elements may be substantially identical in antenna pattern shape and may have already been found to generate a suitable MIMO mapping whatever the configuration originally used. In other words, the serving radio may have multiple radio chains, and each chain may already be mapped to one or more of the individual radiating elements such that the pattern is filled with different radiating elements going to different chains in a many-to-one fashion as the network is known to function.In that case, when transitioning to a different antenna set, it may be desirable for the new set to retain a similar antenna radiator-to-chain mapping so that the MIMO channel performance may be similar. The similarity of the MIMO channel response may itself be derived from a metric function, such as a function that produces a difference metric between two channel or sounding matrices. Such a measure of the MIMO response may itself be obtained during a previous active training phase, or may be provided from decoding of a known condition (such as a preamble in reception or one that occurs during the training process). Some embodiments may give priority to handoff reconfiguration that retains such mapping. Some embodiments may implement a strict pattern mapping, as shown in Figures 38A, 38B, and 38C, where a gray pattern for one chain and a black pattern for another chain within various sets of antenna elements 3806 are preserved across different sets. For example, Figure 38A shows a time when a client device 3504 is served by an antenna made up of gray elements in the set of antenna elements on strip 3500 shown as 3806a, Figure 38B shows a later time when client device 3504 is served by an antenna made up of gray elements in the set of antenna elements shown as 3806f, and Figure 38C shows an even later time when client device 3504 is served by an antenna made up of gray elements in the set of antenna elements shown as 3806n. Preservation of this pattern may be strict in some embodiments or may be based on weighting and priority and may then be overwritten as needs require.

[0100] In some embodiments, an inter-radio handoff may be necessary when a client reaches the end of antenna availability for the serving radio with which it is associated. In some embodiments, both handoff-related reconfiguration and more general reconfiguration can use soft inter-radio handoff in systems with less than full radio options in service. In some embodiments, the handoff may be performed by reconfiguring the new radio with the same identifier as the old radio, so that sufficient state is sent from the old radio to the new radio to enable a cutover of service where the old radio stops serving the client and the new radio starts. In some embodiments, the soft inter-radio handoff process may include a list of the security and association state of that client, as well as any necessary networking or forwarding information. Some minimal embodiments may not carry more than the network configuration state of the service, e.g., which network is being served, etc., and may not carry client state, but instead send a service stop message to cause connected clients to recreate their associations. Some specific embodiments for Wi-Fi may require the new radio to assume the BSSID and service information of the old radio at the point of disconnection, but the new radio may not have any client information. In one such embodiment, this may be all that happens. When the new client sends a message to the access point, such as a power save poll, uplink frame, etc., the new radio will determine that the client is not associated and will send a deauthentication or disassociation message to the client, forcing the client to reestablish its state. Another such embodiment may send a broadcast or unicast deauthentication or disassociation message to the client, forcing the client to proactively reestablish its connection. These messages may be sent from the old radio, the new radio, or both.One possible reason to keep the same BSSID and channel could be to encourage clients to return to a service they were previously using. Changing the BSSID in such a reconfiguration may cause some clients to willingly perform a hard disconnect and scan deeply before re-entering the network.

[0101] In some embodiments, further refinements of the antenna printing process may be used to create directivity on the strip without sacrificing flexibility. FIG. 39A, FIG. 39B show the use of a particular three-dimensional antenna shape 3910 on the strip 3900 to create directivity for the radiation pattern 3905. In some embodiments, the three-dimensional shape may be created using thermal deformation of a plastic underlay, such as thermoforming of a plastic strip. In some embodiments, the plastic may be deformed before printing and a conductive layer may be added on top. In some embodiments, the conductive layer may be stretchable (such as flexible elastic conductive polymers, including metal doped polymers and carbon conductive polymers). In some embodiments, the antenna backhaul traces may be flexible, but the antenna may be affixed (e.g., pre-printed or die cut) or printed on the strip after molding. In some embodiments, the three-dimensional projection of the radiation pattern 3905 shown in FIG. 39A may not be molded into the strip, but may be glued or affixed to the strip 3900. A particular shape may be permanently pressed into the strip 3900, or if the strip 3900 is sufficiently elastic, a shape may be pressed into the strip 3900 at a later time up to and including during deployment. For example, a solid object having a desired shape may be pressed into the antenna portion as a support for the elastic surface. In some embodiments, the antenna may not be printed into the shaped or elastic area, but may be surface bonded to the substrate at one or more points. In some embodiments, the antenna lead itself may be a flexible wire, such as a coaxial cable or a non-radiating cable. In some embodiments, the cable may be set along the elastic area to allow sufficient preloading without binding or breaking, by gluing or bonding at strategic points rather than along its entire length. In one such embodiment, the radiator and leads may be bonded to a rubber sheet that may later be deformed to reorient the radiator.In some embodiments, the flexible, inelastic radiator and leads may be sandwiched between two elastic surfaces so that the antenna cannot bend away from the surfaces and snag or tear. The sandwich itself may be filled with a neutral liquid medium, such as oil, to lubricate the flexible material.

[0102] In some embodiments, the antenna lead itself acts as the antenna. In one such embodiment, a long antenna lead may be sandwiched between ground planes, with ground planes or holes etched into the planes at one or more locations to force that portion of the lead to be radiative. In this manner, a PCB or multi-layer, flat equivalent of a leaky coaxial cable can be constructed.

[0103] The switching network between the RF chains and the antennas is optional. For example, FIG. 40 shows an exemplary embodiment of a 4-way switch fabric 4000 in which no radio can be disconnected from the antenna. In the illustrated example, the switch fabric includes multiple 4-way switches 4004 that connect various RF chains 4002 to antennas 4006. FIG. 41 shows a full matrix switch 4100 of single-pole, single-throw switches 4104 that can couple any combination of RF chains 4102 and antennas 4106, leading to a fully integrated all-to-all map.

[0104] These switching networks may be employed by some embodiments so that multiple chains have options. In various embodiments, blocking, partial blocking, or non-blocking hierarchical switching, including Clos networks of switch banks, may be used to establish the one-to-one mapping property. In some embodiments, the switch banks may be in a multi-way arrangement rather than a row and column arrangement, and thus may generate only permutations. In some embodiments, these permutation networks may be combined with a mixing pool, for example, where more than two leads meet, so that the network can perform many-to-many mappings by partially merging the permutations. These networks may be designed to be sparser than full matrix switches, such as based on design criteria that expect most mappings to be a coherent subset of the available antenna configurations. Note that the RF chains do not have to be on the same synchronized radio. For example, in FIG. 40, FIG. 41, the first three RF chains may be on one radio and the second three RF chains may be on another radio, thus providing a scheme for both radios to potentially share or allocate the shared antenna resources. Each of these embodiments may have different efficiencies (such as the number of switches required to perform a particular task) and blocking parameters.

[0105] Some embodiments include radios of different radio types. These radios may wish to share the same antenna resources. For example, a 2.4 GHz antenna may be useful for both Bluetooth and Wi-Fi. A multi-band antenna may have two or more radiator arms to produce good response at 2.4 GHz and 3.5 GHz, or 5 GHz and 6 GHz. In these embodiments, the different radios may or may not be colliding. Two radios are colliding if one can transmit and the other can receive while both are connected to the same antenna. Thus, the antenna configuration process described above may take into account possible collision rules to generate configurations across multiple radio types, such as between 3.5 GHz CBRS and Wi-Fi. Some embodiments express these constraints as suggestions to a satisfaction engine or as weighted rules in an optimization engine. Some embodiments may use a post-processing step in the discriminator to enforce weighting of values ​​of configurations that result in collisions. In some embodiments, collisions may be expressed as the probability of collision given the expected transmit powers of the two colliding radios. Some embodiments may additionally incorporate inherent configuration characteristics of the antennas and switching paths to the shared antenna, such as additional filter or switch characteristics. In some embodiments, two radios that would otherwise collide are explicitly set aside, at least in part, to avoid collisions, such as by ensuring that appropriate cutoff filters (low pass, high pass, notch, etc.) are placed along the switching fabric, or on the radiators themselves, or at the radio chain input / output ports, to ensure reasonable simultaneous reception and transmission. In some embodiments, the radios may themselves be tuned to ensure pure transmit or receive only bursts, thus enabling simultaneous operation on different bands by avoiding out-of-band roll-off of transmit from interfering with receive.For example, two Wi-Fi radios with similar burst requirements in aggregated transmissions may be scheduled by requesting that the PHY or MAC queues pause while the queues are developed deep enough to ensure the possibility of burst operation. These may be used for multicast / broadcast traffic, or for a mix of unicast traffic on one radio and multicast on another so that both transmitters are turned off before reception is required by the protocol. In some embodiments, the PHYs or queues may be unfrozen at the same time to allow PHY aggregation to take place. In some embodiments, the allowance of simultaneous transmissions and receptions may be further adjusted based on the likelihood of possible collisions. For example, 5 GHz and 6 GHz Wi-Fi radio transmissions may be allowed to proceed uncoordinated if placed on channels far from each other. However, if they are placed on adjacent channels within their roll-off interference distance, the configuration manager may request coordination. In some embodiments, different radios or driving CPUs may possess a shared real-time clock, such as a shared clock or GPIO pulse line, and timestamps that have previously been tightly synchronized. In one exemplary embodiment, the first CPU may send a proposed timestamp to the other CPU to be adopted after the next pulse, and both CPUs may synchronize to the proposed timestamp upon receipt of the next pulse. In some embodiments, predictable interrupt latencies may also be taken into account. In some embodiments, the coherence region may be set based on the collision characteristics of the antennas and associated radios, such that two radios are prohibited (or discouraged as previously discussed) from sharing a coherent set.

[0106] It should be noted that throughout this disclosure, for convenience, switches are described as on / off. However, switches with intermediate settings are available, such as switches designed to operate in the linear region of a FET, or as a tree of binary switches with each leg having a different loss. If variable or adjustable switches are used, the telemetry collected may need to include the settings of each switch. In some embodiments that actively adjust the switches for a learning phase, the switches may always be set to a known value, or to a maximum value. However, a fixed loss corresponding to a particular partial antenna switch setting can be added to the received signal strength to provide an estimate of the value at maximum strength, with the main potential problem being the inability to receive signals below the receive threshold that would have been received if the switch had been at its full setting. Overall, these embodiments may provide a very inexpensive way to generate directional, diverse patterns on an otherwise flat strip.

[0107] According to one embodiment, the techniques described herein are implemented by at least one computing device. The techniques may be implemented in whole or in part using a combination of at least one server computer and / or other computing devices coupled using a network, such as a packet data network. The computing device may be hardwired to perform the techniques, may include digital electronic devices such as at least one application specific integrated circuit (ASIC) or field programmable gate array (FPGA) persistently programmed to perform the techniques, or may include at least one general-purpose hardware processor programmed to perform the techniques according to program instructions in firmware, memory, other storage, or a combination thereof. Such computing devices may also combine custom hardwired logic, ASICs, or FPGAs with custom programming to achieve the described techniques. The computing device may be a server computer, a workstation, a personal computer, a portable computing system, a handheld device, a mobile computing device, a wearable device, a body-worn or implantable device, a smartphone, a smart appliance, an internetworking device, an autonomous or semi-autonomous device such as a robot or an unmanned ground or air vehicle, any other electronic device embedded with hardwired and / or program logic to implement the described techniques, one or more virtual machines or instances in a data center, and / or a network of server computers and / or personal computers.

[0108] Figure 42 is a block diagram illustrating an exemplary computer system on which an embodiment may be implemented. In the example of Figure 42, a computer system 4200 and instructions for implementing the disclosed techniques in hardware, software, or a combination of hardware and software are represented diagrammatically, e.g., as boxes and circles, at the same level of detail that one of ordinary skill in the art to which this disclosure pertains typically uses to communicate computer architecture and computer system implementations.

[0109] The computer system 4200 includes an input / output (I / O) subsystem 4202, which may include a bus and / or other communication mechanism(s) for communicating information and / or instructions between components of the computer system 4200 via electronic signal paths. The I / O subsystem 4202 may include an I / O controller, a memory controller, and at least one I / O port. The electronic signal paths are represented in the drawings diagrammatically, for example, as lines, one-way arrows, or two-way arrows.

[0110] At least one hardware processor 4204 is coupled to the I / O subsystem 4202 for processing information and instructions. The hardware processor 4204 may include, for example, a general-purpose microprocessor or microcontroller and / or an embedded system or special purpose microprocessor such as a graphics processing unit (GPU) or a digital signal processor or an ARM processor. The processor 4204 may include an integrated arithmetic logic unit (ALU) or may be coupled to a separate ALU.

[0111] The computer system 4200 includes one or more units of memory 4206, such as a main memory, coupled to the I / O subsystem 4202 for electronically and digitally storing data and instructions executed by the processor 4204. The memory 4206 may include volatile memory, such as various forms of random access memory (RAM) or other dynamic storage devices. The memory 4206 may also be used to store temporary variables or other intermediate information during execution of instructions to be executed by the processor 4204. Such instructions, when stored on a non-transitory computer-readable storage medium accessible to the processor 4204, can render the computer system 4200 into a special-purpose machine customized to perform the operations specified in the instructions.

[0112] The computer system 4200 further includes a non-volatile memory, such as a read only memory (ROM) 4208 or other static storage device, coupled to the I / O subsystem 4202 for storing information and instructions for the processor 4204. The ROM 4208 may include various forms of programmable ROM (PROM), such as erasable programmable read only memory (EPROM) or electrically erasable programmable read only memory (EEPROM). A unit of persistent storage 4210 may include various forms of non-volatile RAM (NVRAM), such as FLASH memory, or solid-state storage, such as CD-ROM or DVD-ROM, magnetic disks, or optical disks, and may be coupled to the I / O subsystem 4202 for storing information and instructions. The storage 4210 is an example of a non-transitory computer-readable medium that may be used to store instructions and data that, when executed by the processor 4204, cause computer-implemented methods to perform the techniques herein.

[0113] The instructions in memory 4206, ROM 4208 or storage 4210 may comprise one or more sets of instructions organized as modules, methods, objects, functions, routines, or calls. The instructions may be organized as one or more computer programs, operating system services, or application programs, including mobile apps. The instructions may include operating system and / or system software, one or more libraries supporting multimedia, programming, or other functionality, data protocol instructions or stacks implementing TCP / IP, HTTP, or other communications protocols, file format processing instructions for parsing or rendering files encoded using HTML, XML, JPEG, MPEG, or PNG, user interface instructions for rendering or interpreting commands for a graphical user interface (GUI), command line interface, or text user interface, application software such as an office suite, Internet access applications, design and manufacturing applications, graphics applications, audio applications, software engineering applications, educational applications, games, or other applications. The instructions may implement a web server, a web application server, or a web client. The instructions may be organized as a presentation layer, an application layer, and a data storage layer, such as a relational database system with or without Structured Query Language (SQL), an object store, a graph database, a flat file system, or other data storage.

[0114] The computer system 4200 may be coupled to at least one output device 4212 via the I / O subsystem 4202. In one embodiment, the output device 4212 is a digital computer display. Examples of displays that may be used in various embodiments include a touch screen display or a light emitting diode (LED) display or a liquid crystal display (LCD) or an electronic paper display. The computer system 4200 may include other type(s) of output device 4212 instead of or in addition to a display device. Examples of other output devices 4212 include a printer, a ticket printer, a plotter, a projector, a sound or video card, a speaker, a buzzer or a piezoelectric device or other audible device, a lamp or LED or LCD indicator, a tactile device, an actuator or servo.

[0115] At least one input device 4214 is coupled to the I / O subsystem 4202 for communicating signals, data, command selections, or gestures to the processor 4204. Examples of input devices 4214 include touch screens, microphones, still and video digital cameras, alphanumeric and other keys, keypads, keyboards, graphics tablets, image scanners, joysticks, clocks, switches, buttons, dials, slides, and / or various types of sensors, such as force sensors, motion sensors, thermal sensors, accelerometers, gyroscopes, and inertial measurement unit (IMU) sensors, and / or various types of transceivers, such as wireless, radio frequency (RF) or infrared (IR) transceivers, such as cellular or Wi-Fi, and global positioning system (GPS) transceivers.

[0116] Another type of input device is a control device 4216, which can perform other automatic control functions, such as cursor control or navigation in a graphical interface on a display screen, instead of or in addition to input functions. The control device 4216 may be a touchpad, mouse, trackball, or cursor direction keys for communicating directional information and command selections to the processor 4204 and controlling cursor movement on the display 4212. The input device may have at least two degrees of freedom in two axes, a first axis (e.g., x) and a second axis (e.g., y), allowing the device to specify a position in a plane. Another type of input device is a wired, wireless, or optical control device, such as a joystick, wand, console, steering wheel, pedals, gear shift mechanism, etc. The input device 4214 may include a combination of multiple different input devices, such as a video camera and a depth sensor.

[0117] In another embodiment, the computer system 4200 may comprise an Internet of Things (IoT) device in which one or more of the output device 4212, the input device 4214, and the control device 4216 are omitted. Or, in such an embodiment, the input device 4214 may comprise one or more cameras, motion detectors, thermometers, microphones, earthquake detectors, other sensors or detectors, measurement devices or encoders, and the output device 4212 may comprise a special purpose display, such as a single row LED or LCD display, one or more indicators, display panels, meters, valves, solenoids, actuators or servos.

[0118] If the computer system 4200 is a mobile computing device, the input device 4214 may include a Global Positioning System (GPS) receiver coupled to a GPS module that can triangulate to multiple GPS satellites to determine and generate geolocation or position data, such as latitude and longitude values ​​for the physical location of the computer system 4200. The output device 4212 may include hardware, software, firmware, and interfaces for generating location report packets, notifications, pulse or heartbeat signals, or other periodic data transmissions, specifying the location of the computer system 4200, alone or in combination with other application specific data, to the host 4224 or server 4230.

[0119] The computer system 4200 may implement the techniques described herein using customized hardwired logic, at least one ASIC or FPGA, firmware and / or program instructions or logic that, when loaded into the computer system and used or executed in combination, cause or program the computer system to operate as a special purpose machine. According to one embodiment, the techniques described herein are executed by the computer system 4200 in response to the processor 4204 executing at least one sequence of instructions contained in the main memory 4206. Such instructions may be read into the main memory 4206 from another storage medium, such as the storage 4210. Execution of the sequences of instructions contained in the main memory 4206 causes the processor 4204 to perform the process steps described herein. In alternative embodiments, hardwired circuitry may be used in place of or in combination with software instructions.

[0120] The term "storage medium" as used herein refers to any non-transitory medium that stores data and / or instructions that cause a machine to operate in a specific manner. Such storage media may include non-volatile media and / or volatile media. Non-volatile media include, for example, optical or magnetic disks, such as storage 4210. Volatile media include dynamic memory, such as memory 4206. Common forms of storage media include, for example, hard disks, solid state drives, flash drives, magnetic data storage media, any optical or physical data storage media, memory chips, etc.

[0121] Storage media is distinct from but may be used in combination with transmission media, which participate in transferring information between storage media. For example, transmission media include coaxial cables, copper wire and fiber optics, including the wires that comprise a bus in I / O subsystem 4202. Transmission media can also take the form of acoustic or light waves, such as those generated during radio wave and infrared data communications.

[0122] Various forms of media may be involved in carrying the sequence of at least one instruction to the processor 4204 for execution. For example, the instructions may initially be carried on a magnetic disk or solid state drive of a remote computer. The remote computer can load the instructions into its dynamic memory and send the instructions over a communications link, such as a fiber optic or coaxial cable or a telephone line using a modem. A modem or router local to the computer system 4200 can receive the data on the communications link and convert the data into a format readable by the computer system 4200. For example, a receiver, such as a radio frequency antenna or infrared detector, can receive the data carried in a radio or optical signal and appropriate circuitry can provide the data to the I / O subsystem 4202, such as placing the data on a bus. The I / O subsystem 4202 carries the data to memory 4206, from which the processor 4204 retrieves and executes the instructions. The instructions received by memory 4206 may optionally be stored in storage 4210 either before or after execution by the processor 4204.

[0123] The computer system 4200 also includes a communication interface 4218 coupled to the bus 4202. The communication interface 4218 provides a bidirectional data communication coupling to a network link(s) 4220 that is directly or indirectly connected to at least one communication network, such as a network 4222 or a public or private cloud on the Internet. For example, the communication interface 4218 may be an Ethernet network interface, an Integrated Services Digital Network (ISDN) card, a cable modem, a satellite modem, or a modem that provides a data communication connection to a corresponding type of communication line, such as an Ethernet cable or any type of metallic or optical fiber line or a telephone line. The network 4222 broadly represents a local area network (LAN), a wide area network (WAN), a campus network, an internetwork, or any combination thereof. The communication interface 4218 may include a LAN card that provides a data communication connection to a compatible LAN, or a cellular radio telephone interface wired to transmit or receive cellular data according to a cellular radio telephone wireless networking standard, or a satellite wireless interface wired to transmit or receive digital data according to a satellite wireless networking standard. In any such implementation, communication interface 4218 sends and receives electrical, electromagnetic or optical signals over signal paths that carry digital data streams representing various types of information.

[0124] Network link 4220 typically provides electrical, electromagnetic, or optical data communication directly or through at least one network to other data devices using, for example, satellite, cellular, Wi-Fi, or BLUETOOTH technology. For example, network link 4220 may provide a connection through network 4222 to a host computer 4224.

[0125] Further, network link 4220 may provide connectivity through network 4222 or to other computing devices through internetworking equipment and / or computers operated by an Internet Service Provider (ISP) 4226. ISP 4226 provides data communication services through a worldwide packet data communications network represented as Internet 4228. A server computer 4230 may be coupled to Internet 4228. Server 4230 broadly represents any computer, data center, virtual machine, virtual computing instance, or computer running a containerized program system such as DOCKER or KUBERNETES, with or without a hypervisor. Server 4230 may represent an electronic digital service implemented using multiple computers or instances and accessed and used by sending web service requests, uniform resource locator (URL) strings with parameters in an HTTP payload, API calls, app service calls, or other service calls. Computer system 4200 and server 4230 may form elements of a distributed computing system including other computers, processing clusters, server farms, or other organizations of computers that cooperate to perform tasks or run applications or services. The server 4230 may include one or more sets of instructions organized as a module, method, object, function, routine, or call. The instructions may be organized as one or more computer programs, operating system services, or application programs, including mobile apps.The instructions may include operating system and / or system software, one or more libraries supporting multimedia, programming, or other functionality, data protocol instructions or stacks implementing TCP / IP, HTTP, or other communications protocols, file format processing instructions for parsing or rendering files encoded using HTML, XML, JPEG, MPEG, or PNG, user interface instructions for rendering or interpreting commands for a graphical user interface (GUI), a command line interface, or a text user interface, application software such as an office suite, Internet access applications, design and manufacturing applications, graphics applications, audio applications, software engineering applications, educational applications, games, or other applications, etc. Server 4230 may include web application servers hosting a presentation layer, an application layer, and a data storage layer, such as a relational database system using Structured Query Language (SQL) or without SQL, an object store, a graph database, a flat file system, or other data storage.

[0126] Computer system 4200 can send messages and receive data and instructions, including program code, through the network(s), network link 4220 and communication interface 4218. In the Internet example, a server 4230 might transmit a requested code for an application program through the Internet 4228, ISP 4226, local network 4222 and communication interface 4218. The received code may be executed by processor 4204 as it is received, and / or stored in storage 4210, or other non-volatile storage for later execution.

[0127] The execution of instructions described in this section may implement a process consisting of program code and its current activity in the form of an instance of a running computer program. Depending on the operating system (OS), a process may be composed of multiple threads of execution that execute instructions simultaneously. In this context, a computer program is a passive collection of instructions and a process may be the actual execution of those instructions. Several processes may be associated with the same program, e.g., opening multiple instances of the same program often means that multiple processes are running. Multitasking may be implemented to allow multiple processes to share the processor 4204. Although each processor 4204 or processor core executes a single task at a time, the computer system 4200 may be programmed to implement multitasking so that each processor can switch between running tasks without waiting for each task to finish. In one embodiment, the switch may be performed when a task performs an input / output operation, when a task indicates that a task is to be switched, or on a hardware interrupt. Time sharing may be implemented to allow fast response of interactive user applications by performing fast context switches that make it appear as if multiple processes are running simultaneously. In one embodiment, for security and reliability, the operating system may prevent direct communication between independent processes and provide strictly mediated and controlled inter-process communication facilities.

[0128] Throughout this disclosure, the term "Wi-Fi" is used to refer to a particular type of wireless networking based on the IEEE 802.11 standard. However, the technology taught therein applies broadly to wireless networking, and the use of "Wi-Fi" is not to be considered specifically limited to IEEE 802.11 unless clearly stated or implied by the context. Additionally, antenna shapes are illustrated for purposes of connectivity and distinction (such as antennas drawn in different orientations to represent different antenna patterns). These illustrations are conceptual in terms of the shape of the antenna elements, including the ground plane. Specifying specific antenna shapes, applying impedance matching techniques, constructing ground planes, etc., will result in different shapes in actual structures based on manufacturing choices such as materials, thicknesses, conductor weights, etc., and are all well understood in the art without the need for speculation or experimentation.

[0129] In the foregoing specification, embodiments of the invention have been described with reference to numerous specific details that may vary from implementation to implementation. Accordingly, the specification and drawings are to be regarded in an illustrative sense, and not in a restrictive sense. The sole and exclusive indication of the scope of the invention, and what the applicant intends to be the scope of the invention, is the literal and equivalent scope of the set of claims issued from this application, such claims being issued in that particular form, including any subsequent amendments.

[0130] Throughout this disclosure, multiple embodiments are listed that are separate or derived from other embodiments of this disclosure. Furthermore, multiple specific embodiments that may be extensions of more general embodiments are listed throughout this disclosure. It is also understood that combinations and sub-processes of these embodiments are taught by this disclosure, as they can be predicted by those skilled in the art upon, and only upon, reading this disclosure. Furthermore, the use of the plural or singular does not limit the number of items referred to, and unless expressly stated otherwise or logically contradictory, reference to a singular item is to be construed as being in the plural, and vice versa.

[0131] Additionally, multiple alternative embodiments are described throughout this disclosure. Each embodiment differs in tradeoffs or advantages and is thus the best embodiment for that set of tradeoffs and advantages. The choice of which alternative to use depends on the tradeoffs or advantages desired by one of ordinary skill in the art, and such a choice is obvious and straightforward in the art and does not require further invention or discovery. Conditional expressions such as "may," "could," "may" and the like are intended to refer to, and should be construed to be intended to refer to, options (based on manufacture, configuration, or availability) within an embodiment of the invention, and do not state that additional invention is required. For example, a statement that "the invention may respond to a given input" means that one configuration of one assembly of an embodiment of the invention does in fact respond to that input. This is done solely for economy of language and does not imply any uncertainty, incompleteness, or the like regarding the invention taught. This disclosure does not speculate on the future state of the art, but rather describes the current invention. Examples are provided as explicit embodiments of the invention and are provided to elucidate the teachings.

[0132] The present disclosure recites sufficient details to enable one of ordinary skill in the art to construct the surrounding systems and techniques using the novel methods of the invention contained therein without further discovery or invention. [Appendix 1] 1. A computer-implemented method for dynamically configuring a wireless networking system, comprising: obtaining topology information for each of a plurality of antennas in the wireless networking system indicating whether and how the antenna is reaching one or more other antennas in the plurality of antennas, the wireless networking system including two or more Wi-Fi transceivers coupled in parallel to each other, each of which is operable to be selectively coupled to one or more of the plurality of antennas by a respective switch; determining a desired configuration change to the wireless networking system dependent on the obtained topology information, the desired configuration change including a change in which and how many of the plurality of antennas are coupled to the two or more Wi-Fi transceivers or a change in which and how many of the two or more Wi-Fi transceivers are active; initiating the desired configuration change; The method includes: [Appendix 2] initiating the desired configuration change, initiating activation of a switch to couple one of the plurality of antennas to one of the two or more Wi-Fi transceivers; initiating a deactivation of a switch to disconnect one of the plurality of antennas from one of the two or more Wi-Fi transceivers; 2. The method of claim 1, comprising one or more of the following: [Appendix 3] initiating the desired configuration change, initiating activation of one of the two or more Wi-Fi transceivers; initiating a deactivation of one of the two or more Wi-Fi transceivers; 2. The method of claim 1, comprising one or more of the following: [Appendix 4] 2. The method of claim 1, further comprising, for each of the plurality of antennas, populating a respective neighbor table with the obtained topology information. [Appendix 5] The method of claim 1, wherein obtaining the topology information includes using passive or active telemetry to obtain the topology information or neighborhood information from which the topology information is derived during operation of the wireless networking system. [Appendix 6] 6. The method of claim 5, wherein the obtained proximity information includes respective signal strengths associated with transmissions in the wireless networking system. [Appendix 7] 2. The method of claim 1, wherein the obtained topology information is applied to a combination of antennas from which a transmission is received or a combination of receiving antennas. [Appendix 8] 2. The method of claim 1, wherein obtaining the topology information includes correlating transmissions of a transmitting wireless client device among two or more receiving wireless client devices. [Appendix 9] 2. The method of claim 1, wherein determining the desired configuration change is further dependent on respective coverage patterns for wireless transmission and reception for one or more of the plurality of antennas. [Appendix 10] 2. The method of claim 1, wherein determining the desired configuration change includes determining a set of antennas that, when activated, create a coverage pattern over a collective coverage area that includes at least the desired coverage area. [Appendix 11] 11. The method of claim 10, further comprising automatically pruning or shaping the collective coverage area depending on cell location, known interference patterns, or known amounts of directionality of antennas. [Appendix 12] 2. The method of claim 1, wherein determining the desired configuration change further depends on at least one of optimality, network needs, operator requirements, radio conditions, and client requirements. [Appendix 13] an antenna coupled to each Wi-Fi transmitter represents a repeating set of antenna elements that contributes to a default coverage pattern for the set of antenna elements; 2. The method of claim 1, wherein determining the desired configuration change includes determining a configuration change that will result in a change in the coverage pattern for the antenna coupled to one of the two or more Wi-Fi transceivers. [Appendix 14] The wireless networking system comprises: A single networking line and a single power line, and Combined Power and Networking Distribution, At least one of the following: 2. The method of claim 1, wherein the single networking line and the single power line or the combined power and networking distribution are arranged along a linear array and coupled to each of the two or more Wi-Fi transceivers. [Appendix 15] 2. The method of claim 1, wherein determining the desired configuration changes includes using a machine learning model trained to generate a classifier for evaluating proposed configurations for the wireless networking system. [Appendix 16] Furthermore, generating, by the machine learning model, one or more proposed configurations for the wireless networking system; predicting the outcome of each of the proposed configurations; and selecting one of the proposed configurations based on the prediction, the selected proposed configuration including the desired configuration change. [Appendix 17] 16. The method of claim 15, further comprising determining, by the machine learning model, a learned response indicating an importance of each of the plurality of antennas. [Appendix 18] 16. The method of claim 15, wherein the machine learning model is trained using weighted configuration proposals generated by a telemetry process and respective quality measures associated with areas of the wireless networking system affected by each weighted configuration proposal. [Appendix 19] The method further includes determining that the wireless client device should be handed off from one or more currently active antennas through which signals are transmitted or received on behalf of the wireless client device in a first coverage area; the steps of determining the desired configuration change and initiating the desired configuration change are performed in response to a determination that the wireless client device should be handed off from the one or more currently active antennas; 2. The method of claim 1, wherein initiating the desired configuration change includes deactivating at least one of the one or more currently active antennas and activating at least one alternative antenna having a second coverage area. [Appendix 20] 20. The method of claim 19, wherein determining that the wireless client device should be handed off from the one or more currently active antennas includes determining that the wireless client device is approaching a boundary of the first coverage area. [Appendix 21] The method further includes obtaining, during operation of the wireless networking system, a measure of a signal quality, an error rate, or a data rate associated with the wireless client device and the first Wi-Fi transceiver, or a measure of a signal quality, an error rate, or a data rate associated with the wireless client device and the second Wi-Fi transceiver, using passive or active telemetry; 20. The method of claim 19, wherein determining that the wireless client device should be handed off from the one or more currently active antennas is dependent on a signal quality, error rate, or data rate associated with the wireless client device and the first Wi-Fi transceiver, or a signal quality, error rate, or data rate associated with the wireless client device and the second Wi-Fi transceiver. [Appendix 22] The step of determining a desired configuration comprises: modifying the configuration suggestions or the weighting of the configuration suggestions; initiating generation of a new configuration for the wireless networking system dependent on the revised configuration proposals or weightings of configuration proposals; 20. The method of claim 19, further comprising generating a new configuration for the wireless networking system depending on the revised configuration proposals or weightings of configuration proposals and one or more applicable configuration change constraints. [Appendix 23] The one or more applicable configuration change constraints: a constraint on the distance between the first coverage area and the second coverage area; and a requirement that the at least one alternative antenna be in the same coherent antenna set as the one or more currently active antennas; or 23. The method of claim 22 including a requirement that the at least one alternate antenna represent a same subset of elements in a first instance of a repeating set of antenna elements and a same subset of elements in a second instance of the repeating set of antenna elements represented by the one or more currently active antennas. [Appendix 24] One or more computer-readable non-transitory storage media storing one or more sequences of instructions that, when executed by one or more processors, cause the one or more processors to: obtaining topology information for each of a plurality of antennas in a wireless networking system indicative of whether and how the antenna is reaching one or more other antennas in the plurality of antennas, the wireless networking system including two or more Wi-Fi transceivers coupled in parallel with each other and each operable to be selectively coupled to one or more of the plurality of antennas by a respective switch; determining a desired configuration change to the wireless networking system dependent on the obtained topology information, the desired configuration change including a change in which and how many of the plurality of antennas are coupled to the two or more Wi-Fi transceivers, or a change in which antennas or how many of the two or more Wi-Fi transceivers are active; One or more computer-readable non-transitory storage media that initiate the desired configuration changes. [Appendix 25] 25. The medium of claim 24, wherein to obtain the topology information, the instructions cause the one or more processors to use passive or active telemetry to obtain the topology information or neighborhood information from which the topology information is derived during operation of the wireless networking system. [Appendix 26] 25. The medium of claim 24, wherein the instructions cause the one or more processors to correlate transmissions of a transmitting wireless client device between two or more receiving wireless client devices to obtain the topology information. [Appendix 27] 25. The medium of claim 24, wherein determining the desired configuration change is further dependent on respective coverage patterns for wireless transmission and reception for one or more of the plurality of antennas. [Appendix 28] 25. The medium of claim 24, wherein the instructions cause the one or more processors to determine a set of antennas that, when activated, creates a coverage pattern over a collective coverage area that includes at least a desired coverage area, to determine the desired configuration change. [Appendix 29] the antennas coupled to each Wi-Fi transmitter represent a repeating collection of antenna elements, each of which contributes to a default coverage pattern for the collection of antenna elements; 25. The medium of claim 24, wherein the instructions to determine the desired configuration change cause the one or more processors to determine a configuration change that will result in a change in the coverage pattern for the antenna coupled to one of the two or more Wi-Fi transceivers. [Appendix 30] 25. The medium of claim 24, wherein the instructions cause the one or more processors to use a trained machine learning model to generate a classifier for evaluating proposed configurations for the wireless networking system to determine the desired configuration changes. [Appendix 31] The instructions further include causing the one or more processors to: generating one or more proposed configurations for the wireless networking system with the machine learning model; predicting the outcome of each of the proposed configurations; 31. The medium of claim 30, further comprising: selecting one of the proposed configurations based on the prediction, the selected proposed configuration including the desired configuration change. [Appendix 32] 31. The medium of claim 30, wherein the machine learning model is trained using weighted configuration proposals generated by a telemetry process and respective quality measures associated with areas of the wireless networking system affected by each weighted configuration proposal. [Appendix 33] The instructions further cause the one or more processors to determine that the wireless client device should be handed off from one or more currently active antennas through which signals are transmitted or received on behalf of the wireless client device in a first coverage area; determining the desired configuration change and initiating the desired configuration change are performed in response to determining that the wireless client device should be handed off from the one or more currently active antennas; 25. The method of claim 24, wherein to perform the initiation of the desired configuration change, the instructions cause the one or more processors to deactivate at least one of the one or more currently active antennas and activate at least one alternate antenna having a second coverage area. [Appendix 34] 34. The medium of claim 33, wherein the instructions cause the one or more processors to determine that the wireless client device should be handed off from the one or more currently active antennas. [Appendix 35] The instructions further cause the one or more processors to obtain, during operation of the wireless networking system, measures of signal quality, error rates, or data rates associated with the wireless client device and the first Wi-Fi transceiver, or measures of signal quality, error rates, or data rates associated with the wireless client device and a second Wi-Fi transceiver, using passive or active telemetry; determining that the wireless client device should be handed off from the one or more currently active antennas is dependent on a signal quality, an error rate, or a data rate associated with the wireless client device and the first Wi-Fi transceiver, or a signal quality, an error rate, or a data rate associated with the wireless client device and the second Wi-Fi transceiver; 33. The medium described in Appendix 33. [Appendix 36] To determine the desired configuration, the instructions cause the one or more processors to: Modify the configuration proposals or the weighting of the configuration proposals; Initiating the creation of a new configuration of the wireless networking system dependent on the revised configuration proposals or weightings of the configuration proposals; 34. The medium of claim 33, wherein the instructions further cause the one or more processors to generate the new configuration for the wireless networking system depending on the revised configuration proposals or weightings of configuration proposals and one or more applicable configuration change constraints. [Appendix 37] The one or more applicable configuration change constraints: a constraint on the distance between the first coverage area and the second coverage area; and a requirement that the at least one alternative antenna be in the same coherent antenna set as the one or more currently active antennas; or and a requirement that the at least one alternate antenna represent a same subset of elements in a first instance of a repeating set of antenna elements and a same subset of elements in a second instance of the repeating set of antenna elements represented by the one or more currently active antennas. [Appendix 38] 1. A computer-implemented method for dynamically configuring a wireless networking system, comprising: obtaining topology information for each of a plurality of antennas in the wireless networking system indicating whether and how the antenna reaches one or more other antennas in the plurality of antennas, the wireless networking system including two or more Wi-Fi transceivers coupled in parallel with each other and each operable to be selectively coupled to one or more of the plurality of antennas by a respective switch, the obtained topology information representing one or more configuration suggestions for the wireless networking system; determining a predicted outcome for each of the one or more configuration proposals depending on the acquired topology information using a trained machine learning model to generate a classifier for evaluating configuration proposals for the wireless networking system; and determining, depending on the prediction result, a set of antennas selected from the plurality of antennas to be activated in the wireless networking system after a configuration change to the wireless networking system, the configuration change including a change in which and how many of the plurality of antennas are coupled to the two or more Wi-Fi transceivers or a change in which or how many of the two or more Wi-Fi transceivers are active. [Appendix 39] 39. The method of claim 38, further comprising determining, by the machine learning model, a learned response indicating an importance of each of the plurality of antennas. [Appendix 40] 39. The method of claim 38, wherein the machine learning model is trained using the one or more configuration proposals weighted by the telemetry process and respective quality measures associated with areas of the wireless networking system affected by each weighted configuration proposal. [Appendix 41] 39. The method of claim 38, wherein obtaining the topology information includes using passive or active telemetry to obtain the topology information, or neighborhood information from which the topology information is derived, during operation of the wireless networking system. [Appendix 42] 39. The method of claim 38, wherein the obtained topology information is applied to an antenna combination from which a transmission is received or to a receiving antenna combination. [Appendix 43] 39. The method of claim 38, wherein determining the set of antennas is further dependent on respective coverage patterns for wireless transmission and reception for one or more of the plurality of antennas. [Appendix 44] 39. The method of claim 38, wherein determining the set of antennas includes determining a set of antennas that, when activated, create a coverage pattern over a collective coverage area that includes at least the desired coverage area. [Appendix 45] 39. The method of claim 38, wherein determining the set of antennas is further dependent on at least one of optimality, network needs, operator requirements, radio conditions, and client requirements. [Appendix 46] 39. The method of claim 38, further comprising initiating the configuration change. [Appendix 47] The method further includes determining that the wireless client device should be handed off from one or more currently active antennas through which signals are transmitted or received on behalf of the wireless client device in a first coverage area; the steps of determining the respective predicted outcomes and determining the set of antennas are performed in response to a determination that the wireless client device should be handed off from the one or more currently active antennas; 47. The method of claim 46, wherein initiating the configuration change includes deactivating at least one of the one or more currently active antennas and activating at least one alternative antenna having a second coverage area. [Appendix 48] 48. The method of claim 47, wherein determining that the wireless client device should be handed off from the one or more currently active antennas includes determining that the wireless client device is approaching a boundary of a first coverage area. [Appendix 49] The method further includes obtaining, during operation of the wireless networking system, a measure of a signal quality, an error rate, or a data rate associated with the wireless client device and the first Wi-Fi transceiver, or a measure of a signal quality, an error rate, or a data rate associated with the wireless client device and the second Wi-Fi transceiver, using passive or active telemetry; 48. The method of claim 47, wherein determining that the wireless client should be handed off from the one or more currently active antennas is dependent on a signal quality, error rate, or data rate associated with the wireless client device and the first Wi-Fi transceiver or a signal quality, error rate, or data rate associated with the wireless client device and the second Wi-Fi transceiver. [Appendix 50] The step of determining a set of antennas comprises: modifying the configuration suggestions or the weighting of the configuration suggestions; and initiating generation of a new configuration for the wireless networking system depending on the revised configuration proposal or a weighting of the configuration proposals; 48. The method of claim 47, further comprising generating the new configuration of the wireless networking system depending on the revised configuration proposals or weightings of configuration proposals and one or more applicable configuration change constraints. [Appendix 51] The one or more applicable configuration change constraints are: a constraint on the distance between the first coverage area and the second coverage area; and a requirement that the at least one alternative antenna be in the same coherent antenna set as the one or more currently active antennas; or 51. The method of claim 50 including a requirement that the at least one alternate antenna represent a same subset of elements in a first instance of a repeating set of antenna elements and a same subset of elements in a second instance of the repeating set of antenna elements represented by the one or more currently active antennas. [Appendix 52] One or more computer-readable non-transitory storage media storing one or more sequences of instructions that, when executed by one or more processors, cause the one or more processors to: obtaining topology information for each of a plurality of antennas in a wireless networking system indicative of whether and how the antenna is reaching one or more other antennas in the plurality of antennas, the wireless networking system including two or more Wi-Fi transceivers coupled in parallel with each other and each operable to be selectively coupled to one or more of the plurality of antennas by a respective switch, the obtained topology information representing one or more configuration proposals for the wireless networking system; determining a predicted outcome for each of the one or more configuration proposals depending on the acquired topology information using the trained machine learning model to generate a classifier for evaluating configuration proposals for the wireless networking system; A computer-readable non-transitory storage medium that determines, depending on the prediction result, a set of antennas selected from the plurality of antennas to be activated in the wireless networking system after a configuration change to the wireless networking system, the configuration change including a change in which and how many of the plurality of antennas are coupled to the two or more Wi-Fi transceivers or a change in which or how many of the two or more Wi-Fi transceivers are active. [Appendix 53] 53. The medium of claim 52, wherein the machine learning model is trained using the one or more configuration proposals weighted by a telemetry process and respective quality measures associated with areas of the wireless networking system affected by each weighted configuration proposal. [Appendix 54] 53. The medium of claim 52, wherein to obtain the topology information, the instructions cause the one or more processors to obtain the topology information, or neighborhood information from which the topology information is derived, during operation of the wireless networking system using passive or active telemetry. [Appendix 55] 53. The medium of claim 52, wherein the step of determining the set of antennas is further dependent upon respective coverage patterns for wireless transmission and reception for one or more antennas of the plurality of antennas. [Appendix 56] 53. The medium of claim 52, wherein the instructions for determining the set of antennas cause the one or more processors to determine a set of antennas that, when activated, creates a coverage pattern over a collective coverage area that includes at least a desired coverage area. [Appendix 57] 53. The medium of claim 52, wherein the instructions further cause the one or more processors to initiate a configuration change. [Appendix 58] The instructions further cause the one or more processors to determine that the wireless client device should be handed off from one or more currently active antennas through which signals are transmitted or received on behalf of the wireless client device in a first coverage area; the steps of determining each predicted outcome and determining the set of antennas are performed in response to determining that the wireless client device should be handed off from the one or more currently active antennas; 53. The medium of claim 52, wherein to initiate the configuration change, the instructions cause the one or more processors to deactivate at least one of the one or more currently active antennas and activate at least one alternate antenna having a second coverage area. [Appendix 59] 59. The medium of claim 58, wherein the instructions cause the one or more processors to determine that the wireless client device should be handed off from the one or more currently active antennas. [Appendix 60] The instructions further cause the one or more processors to obtain, during operation of the wireless networking system, measures of signal quality, error rates, or data rates associated with the wireless client device and the first Wi-Fi transceiver, or measures of signal quality, error rates, or data rates associated with the wireless client device and a second Wi-Fi transceiver, using passive or active telemetry; 59. The medium of claim 58, wherein determining that the wireless client device should be handed off from the one or more currently active antennas is dependent on a signal quality, error rate, or data rate associated with the wireless client device and the first Wi-Fi transceiver or a signal quality, error rate, or data rate associated with the wireless client device and the second Wi-Fi transceiver. [Appendix 61] To determine the set of antennas, the instructions cause the one or more processors to: Modify the configuration proposals or the weighting of the configuration proposals; Initiating the creation of a new configuration of the wireless networking system dependent on the revised configuration proposals or weightings of the configuration proposals; 59. The medium of claim 58, wherein the instructions further cause the one or more processors to generate the new configuration for the wireless networking system depending on the revised configuration proposals or weightings of configuration proposals and one or more applicable configuration change constraints. [Appendix 62] The one or more applicable configuration change constraints: a constraint on the distance between the first coverage area and the second coverage area; and a requirement that the at least one alternative antenna be in the same coherent antenna set as the one or more currently active antennas; or and a requirement that the at least one alternate antenna represent a same subset of elements in a first instance of a repeating set of antenna elements and a same subset of elements in a second instance of the repeating set of antenna elements represented by the one or more currently active antennas. [Appendix 63] 1. A computer-implemented method for performing antenna handoff in a wireless networking system, comprising: determining that a wireless client device should be handed off from one or more currently active antennas through which signals are transmitted or received on behalf of the wireless client device in a first coverage area, the wireless networking system including two or more Wi-Fi transceivers coupled in parallel with one another and each operable to be selectively coupled to one or more of a plurality of antennas by a respective switch, the one or more currently active antennas being coupled to a first Wi-Fi transceiver of the two or more Wi-Fi transceivers; and initiating a configuration change of the wireless networking system, the configuration change including deactivating at least one of the one or more currently active antennas and activating at least one alternative antenna having a second coverage area. [Appendix 64] 64. The method of claim 63, wherein determining that the wireless client device should be handed off from the one or more currently active antennas includes determining that the wireless client device is approaching a boundary of the first coverage area. [Appendix 65] 64. The method of claim 63, wherein the at least one alternate antenna is coupled to the first Wi-Fi transceiver of the two or more Wi-Fi transceivers. [Appendix 66] 64. The method of claim 63, wherein the at least one alternative antenna is coupled to a second Wi-Fi transceiver of the two or more Wi-Fi transceivers. [Appendix 67] 67. The method of claim 66, further comprising the step of deactivating the first Wi-Fi transceiver of the two or more Wi-Fi transceivers. [Appendix 68] 64. The method of claim 63, wherein determining that the wireless client device should be handed off from the one or more currently active antennas is dependent on a signal quality, error rate, or data rate associated with the wireless client device and the first Wi-Fi transceiver, or a signal quality error rate, or data rate associated with the wireless client device and a second Wi-Fi transceiver. [Appendix 69] 69. The method of claim 68, further comprising obtaining measures of the signal quality, error rate, or data rate associated with the wireless client device and the first or second Wi-Fi transceiver during operation of the wireless networking system using passive or active telemetry. [Appendix 70] 64. The method of claim 63, wherein initiating the configuration change includes controlling one or more of the respective switches to deactivate the one or more antennas and activate the at least one alternate antenna. [Appendix 71] The step of initiating the configuration change includes: selecting a first pre-configured set of antennas from among a plurality of pre-configured sets of antennas, the first pre-configured set including the at least one alternative antenna; and controlling one or more of the respective switches to cause a movement of signals transmitted or received on behalf of the wireless client device from a second pre-configured set of antennas that includes the one or more currently active antennas to the first pre-configured set of antennas. [Appendix 72] The step of initiating the configuration change includes: modifying the configuration suggestions or the weighting of the configuration suggestions; and initiating creation of a new configuration of the wireless networking system depending on the revised configuration proposal or weighting of the configuration proposal. [Appendix 73] 73. The method of claim 72, further comprising generating the new configuration of the wireless networking system depending on the revised configuration proposals or weightings of configuration proposals and one or more applicable configuration change constraints. [Appendix 74] 74. The method of claim 73, wherein the one or more applicable configuration change constraints include a constraint on a distance between the first coverage area and the second coverage area. [Appendix 75] the plurality of antennas includes a plurality of coherent antenna sets, each of which includes a plurality of antennas; 74. The method of claim 73, wherein the one or more applicable configuration change constraints include a requirement that the at least one alternative antenna be in a same coherent antenna set as the one or more currently active antennas. [Appendix 76] the plurality of antennas includes repeating sets of antenna elements each of which contributes to a default coverage pattern for the set of antenna elements; the currently active antenna represents a subset of elements in a first instance of the repeating set of antenna elements; 74. The method of claim 73, wherein the one or more applicable configuration change constraints include a requirement that the at least one alternative antenna represent a same subset of elements in a second instance of the repeating set of antenna elements. [Appendix 77] obtaining, for each of the plurality of antennas, topology information indicating whether and how the antenna is within range of one or more other antennas in the plurality of antennas; determining said configuration changes in dependence on said obtained topology information; 64. The method of claim 63, comprising: [Appendix 78] 78. The method of claim 77, wherein obtaining the topology information includes using passive or active telemetry to obtain the topology information or neighborhood information from which the topology information is derived during operation of the wireless networking system. [Appendix 79] 78. The method of claim 77, wherein determining the configuration change includes determining a set of antennas that, when activated, creates a coverage pattern across an aggregate coverage area that includes at least the second coverage area. [Appendix 80] 78. The method of claim 77, wherein the step of determining a configuration change includes using a machine learning model trained to generate a classifier for evaluating proposed configurations of the wireless networking system. [Appendix 81] further comprising generating, by the machine learning model, one or more proposed configurations for the wireless networking system; predicting the outcome of each of the proposed configurations; 81. The method of claim 80, further comprising selecting one of the proposed configurations based on the prediction, the selected proposed configuration including the configuration change. [Appendix 82] 81. The method of claim 80, wherein the machine learning model is trained using weighted configuration proposals generated by a telemetry process and respective quality measures associated with areas of the wireless networking system affected by each weighted configuration proposal. [Appendix 83] One or more computer-readable non-transitory storage media having stored thereon one or more sequences of instructions that, when executed using one or more processors, cause the one or more processors to: causing a wireless networking system to determine that a wireless client device should be handed off from one or more currently active antennas through which signals are transmitted or received on behalf of the wireless client device in a first coverage area, the wireless networking system including two or more Wi-Fi transceivers coupled in parallel with one another and each operable to be selectively coupled to one or more of a plurality of antennas by a respective switch, the one or more currently active antennas coupled to a first Wi-Fi transceiver of the two or more Wi-Fi transceivers; and initiating a configuration change of the wireless networking system, the configuration change including deactivating at least one of the one or more currently active antennas and activating at least one alternative antenna having a second coverage area. [Appendix 84] To determine that the wireless client device should be handed off from one or more currently active antennas, the instructions cause the one or more processors to determine that the wireless client device is approaching a boundary of the first coverage area. 83. The medium described in Appendix 83. [Appendix 85] The instructions further cause the one or more processors to obtain, during operation of the wireless networking system, measures of signal quality, error rates, or data rates associated with the wireless client device and the first Wi-Fi transceiver, or measures of signal quality, error rates, or data rates associated with the wireless client device and a second Wi-Fi transceiver, using passive or active telemetry; 84. The medium of claim 83, wherein determining that the wireless client device should be handed off from one or more currently active antennas is dependent on the signal quality, error rate, or data rate associated with the wireless client device and the first Wi-Fi transceiver or the signal quality, error rate, or data rate associated with the wireless client device and the second Wi-Fi transceiver. [Appendix 86] To initiate the configuration change, the instructions cause the one or more processors to: selecting a first pre-configured set of antennas from among a plurality of pre-configured sets of antennas, the first pre-configured set including the at least one alternative antenna; 84. The medium of claim 83, controlling one or more of the respective switches to effect a movement of signals transmitted or received on behalf of the wireless client device from a second pre-configured set of antennas, including the one or more currently active antennas, to the first pre-configured set of antennas. [Appendix 87] To initiate the configuration change, the instructions cause the one or more processors to: Modify the configuration proposals or the weighting of the configuration proposals; Initiating the creation of a new configuration of the wireless networking system dependent on the revised configuration proposals or weightings of the configuration proposals; 84. The medium of claim 83, generating a new configuration of the wireless networking system dependent on the revised configuration proposals or weightings of configuration proposals and one or more applicable configuration change constraints. [Appendix 88] The one or more applicable configuration change constraints: a constraint on the distance between the first coverage area and the second coverage area; and a requirement that the at least one alternative antenna be in the same coherent antenna set as the one or more currently active antennas; or and a requirement that the at least one alternate antenna represent a same subset of elements in a first instance of a repeating set of antenna elements and a same subset of elements in a second instance of the repeating set of antenna elements represented by the one or more currently active antennas. [Appendix 89] The instructions further include causing the one or more processors to: obtaining, for each of the plurality of antennas, topology information indicative of whether and how the antenna is within range of one or more other antennas in the plurality of antennas; 84. The medium of claim 83, wherein the configuration change is determined depending on the acquired topology information. [Appendix 90] 90. The medium of claim 89, wherein to obtain the topology information, the instructions cause the one or more processors to obtain the topology information or neighborhood information from which the topology information is derived using passive or active telemetry during operation of the wireless networking system. [Appendix 91] 90. The medium of claim 89, wherein the instructions cause the one or more processors to determine a set of antennas that, when activated, creates a coverage pattern over an aggregate coverage area that includes at least the second coverage area, to determine the configuration change. [Appendix 92] 90. The medium of claim 89, wherein the instructions cause the one or more processors to use a trained machine learning model to generate a classifier for evaluating proposed configurations of the wireless networking system to determine the configuration changes. [Appendix 93] The instructions further include causing the one or more processors to: generating, with the machine learning model, one or more proposed configurations for the wireless networking system; predicting the outcome of each of the proposed configurations; 93. The medium of claim 92, further comprising: selecting one of the proposed configurations based on the prediction, the selected proposed configuration including the configuration change. [Appendix 94] The medium of claim 92, wherein the machine learning model is trained using weighted configuration proposals generated by a telemetry process and respective quality measures associated with areas of the wireless networking system affected by each weighted configuration proposal. [Explanation of symbols]

[0133] 100, 200 Network 110 Wireless Light Bulb 120, 290 Wi-Fi Transceiver 130, 295 Antenna 210 Main power supply 220 Power supply 230 Power Line Network Transceiver 250 CPUs / Controllers 260 Access Point Manager 270 Service Provision 280 List

Claims

1. 1. A computer-implemented method for dynamically configuring a wireless networking system, comprising: obtaining topology information for each of a plurality of antennas in the wireless networking system indicating whether and how a signal of each of the plurality of antennas is reaching one or more other antennas in the plurality of antennas, the wireless networking system including two or more Wi-Fi transceivers coupled to each other in a parallel configuration when coupled to a single networking line and a single power line arranged along a linear array, each of the two or more Wi-Fi transceivers being selectively coupled to one or more of the plurality of antennas by a respective switch; determining a desired configuration change to the wireless networking system dependent on the obtained topology information, the desired configuration change including a change in which and how many of the plurality of antennas are coupled to the two or more Wi-Fi transceivers or a change in which and how many of the two or more Wi-Fi transceivers are active; initiating the desired configuration change; The method includes:

2. initiating the desired configuration change, initiating activation of a switch to couple one of the plurality of antennas to one of the two or more Wi-Fi transceivers; initiating a deactivation of a switch to disconnect one of the plurality of antennas from one of the two or more Wi-Fi transceivers; Including one or more of the following: Or, initiating activation of one of the two or more Wi-Fi transceivers; initiating a deactivation of one of the two or more Wi-Fi transceivers; The method of claim 1 , comprising one or more of:

3. The method of claim 1 , further comprising the step of: for each of the plurality of antennas, populating a respective neighbor table with the obtained topology information.

4. The obtained topology information is applied to a combination of antennas on which a transmission is received or to a combination of receiving antennas; Or, obtaining the topology information includes correlating transmissions of a transmitting wireless client device among two or more receiving wireless client devices; Or, obtaining the topology information includes obtaining the topology information or neighborhood information from which the topology information is derived during operation of the wireless networking system using passive or active telemetry; Optionally, The method of claim 1 , wherein the obtained proximity information includes respective signal strengths associated with transmissions in the wireless networking system.

5. The step of determining the desired configuration change further comprises: dependent on respective coverage patterns for radio transmission and reception for one or more of the plurality of antennas; Or, The method of claim 1 , wherein the optimality depends on at least one of network needs, operator requirements, radio conditions, and client requirements.

6. determining the desired configuration change includes determining a set of antennas that, when activated, creates a coverage pattern over an aggregate coverage area that includes at least the desired coverage area; Optionally, The method of claim 1 , further comprising automatically pruning or shaping the collective coverage area depending on cell location, known interference patterns, or known amounts of directionality of antennas.

7. an antenna coupled to each Wi-Fi transmitter represents a repeating set of antenna elements that contributes to a default coverage pattern for the set of antenna elements; 2. The method of claim 1 , wherein determining the desired configuration change comprises determining a configuration change that results in a change in the coverage pattern for the antenna coupled to one of the two or more Wi-Fi transceivers.

8. The wireless networking system comprises: the single networking line and the single power line; and Combined Power and Networking Distribution, At least one of the following:

2. The method of claim 1, wherein the single networking line and the single power line or the combined power and networking distribution are arranged along the linear array and coupled to each of the two or more Wi-Fi transceivers.

9. determining the desired configuration change includes using a trained machine learning model to generate a classifier for evaluating proposed configurations for the wireless networking system; Optionally, The method further comprises: generating, by the machine learning model, one or more proposed configurations for the wireless networking system; predicting the outcome of each of the proposed configurations; selecting one of the proposed configurations based on the prediction, the selected proposed configuration including the desired configuration change; Or, The method further includes determining, by the machine learning model, a learned response indicative of an importance of each of the plurality of antennas; Or, 2. The method of claim 1, wherein the machine learning model is trained using weighted configuration proposals generated by a telemetry process and respective quality measures associated with areas of the wireless networking system affected by each weighted configuration proposal.

10. The method further includes determining that the wireless client device should be handed off from one or more currently active antennas through which signals are transmitted or received on behalf of the wireless client device in a first coverage area; the steps of determining the desired configuration change and initiating the desired configuration change are performed in response to a determination that the wireless client device should be handed off from the one or more currently active antennas; Initiating the desired configuration change includes deactivating at least one of the one or more currently active antennas and activating at least one alternative antenna having a second coverage area; Optionally, determining that the wireless client device should be handed off from the one or more currently active antennas includes determining that the wireless client device is approaching a boundary of the first coverage area; Or, The method further comprises: obtaining, during operation of the wireless networking system, a measure of a signal quality, an error rate, or a data rate associated with the wireless client device and the first Wi-Fi transceiver, or a measure of a signal quality, an error rate, or a data rate associated with the wireless client device and the second Wi-Fi transceiver, using passive or active telemetry; determining that the wireless client device should be handed off from the one or more currently active antennas is dependent on a signal quality, an error rate, or a data rate associated with the wireless client device and the first Wi-Fi transceiver, or a signal quality, an error rate, or a data rate associated with the wireless client device and the second Wi-Fi transceiver; Or, The step of determining a desired configuration comprises: modifying the configuration suggestions or the weighting of the configuration suggestions; initiating generation of a new configuration for the wireless networking system dependent on the revised configuration proposals or weightings of configuration proposals; The method further includes generating a new configuration for the wireless networking system depending on the revised configuration proposals or weightings of configuration proposals and one or more applicable configuration change constraints; Optionally, The one or more applicable configuration change constraints: a constraint on the distance between the first coverage area and the second coverage area; and a requirement that the at least one alternative antenna be in the same coherent antenna set as the one or more currently active antennas; or 2. The method of claim 1 , including a requirement that the at least one alternate antenna represent a subset of elements in a first instance of a repeating set of antenna elements and a subset of elements in a second instance of the repeating set of antenna elements represented by the one or more currently active antennas.

11. One or more computer-readable non-transitory storage media storing one or more sequences of instructions that, when executed by one or more processors, cause the one or more processors to: obtaining topology information for each of a plurality of antennas in a wireless networking system indicative of whether and how a signal of each of the plurality of antennas is reaching one or more other antennas in the plurality of antennas, the wireless networking system including two or more Wi-Fi transceivers coupled to each other in a parallel configuration when coupled to a single networking line and a single power line arranged along a linear array, each of the two or more Wi-Fi transceivers selectively coupled to one or more of the plurality of antennas by a respective switch; determining a desired configuration change to the wireless networking system dependent on the obtained topology information, the desired configuration change including a change in which and how many of the plurality of antennas are coupled to the two or more Wi-Fi transceivers, or a change in which antennas or how many of the two or more Wi-Fi transceivers are active; One or more computer-readable non-transitory storage media for initiating the desired configuration changes.

12. To obtain the topology information, the instructions cause the one or more processors to use passive or active telemetry to obtain the topology information or neighborhood information from which the topology information is derived during operation of the wireless networking system; Or, 12. The medium of claim 11, wherein the instructions cause the one or more processors to correlate transmissions of a transmitting wireless client device between two or more receiving wireless client devices to obtain the topology information.

13. determining the desired configuration change further comprises depending on respective coverage patterns for radio transmission and reception for one or more antennas of the plurality of antennas; Or, To determine the desired configuration change, the instructions cause the one or more processors to determine a set of antennas that, when activated, creates a coverage pattern over a collective coverage area that includes at least a desired coverage area; Or, the antennas coupled to each Wi-Fi transmitter represent a repeating collection of antenna elements, each of which contributes to a default coverage pattern for the collection of antenna elements; 12. The medium of claim 11, wherein to determine the desired configuration change, the instructions cause the one or more processors to determine a configuration change that will result in a change in the coverage pattern for the antenna coupled to one of the two or more Wi-Fi transceivers.

14. To determine the desired configuration change, the instructions cause the one or more processors to use a trained machine learning model to generate a classifier for evaluating proposed configurations for the wireless networking system; Optionally, The instructions further include causing the one or more processors to: generating one or more proposed configurations for the wireless networking system with the machine learning model; predicting the outcome of each of the proposed configurations; selecting one of the proposed configurations based on the prediction, the selected proposed configuration including the desired configuration change; Or, 12. The medium of claim 11, wherein the machine learning model is trained using weighted configuration proposals generated by a telemetry process and respective quality measures associated with areas of the wireless networking system affected by each weighted configuration proposal.

15. The instructions further cause the one or more processors to determine that the wireless client device should be handed off from one or more currently active antennas through which signals are transmitted or received on behalf of the wireless client device in a first coverage area; determining the desired configuration change and initiating the desired configuration change are performed in response to determining that the wireless client device should be handed off from the one or more currently active antennas; To perform the initiation of the desired configuration change, the instructions cause the one or more processors to deactivate at least one of the one or more currently active antennas and activate at least one alternate antenna having a second coverage area; Optionally, To determine that the wireless client device should be handed off from the one or more currently active antennas, the instructions cause the one or more processors to determine that the wireless client device is approaching a boundary of the first coverage area; Or, The instructions further cause the one or more processors to obtain, during operation of the wireless networking system, measures of signal quality, error rates, or data rates associated with the wireless client device and the first Wi-Fi transceiver, or measures of signal quality, error rates, or data rates associated with the wireless client device and a second Wi-Fi transceiver, using passive or active telemetry; determining that the wireless client device should be handed off from the one or more currently active antennas is dependent on a signal quality, an error rate, or a data rate associated with the wireless client device and the first Wi-Fi transceiver, or a signal quality, an error rate, or a data rate associated with the wireless client device and the second Wi-Fi transceiver; Or, To determine the desired configuration, the instructions cause the one or more processors to: Modify the configuration proposals or the weighting of the configuration proposals; Initiating the creation of a new configuration of the wireless networking system dependent on the revised configuration proposals or weightings of the configuration proposals; The instructions further cause the one or more processors to generate the new configuration for the wireless networking system depending on the revised configuration proposal or weightings of configuration proposals and one or more applicable configuration change constraints. Optionally, The one or more applicable configuration change constraints: a constraint on the distance between the first coverage area and the second coverage area; and a requirement that the at least one alternative antenna be in the same coherent antenna set as the one or more currently active antennas; or and a requirement that the at least one alternate antenna represent a subset of elements in a first instance of a repeating set of antenna elements that is the same as a subset of elements in a second instance of the repeating set of antenna elements represented by the one or more currently active antennas.

Images