Wireless communication networks and transceivers

Abstract

A wireless communication network and a transceiver are provided. A radio transmitting device is configured to transmit a spread spectrum radio signal in which a carrier frequency is changed among a predetermined set of radio channels according to a hopping sequence, the radio signal being organized into packets, each packet having a header transmitted at a first channel in the hopping sequence, the header including a detection sequence, and payload data encoding a message transmitted at a subsequent channel in the hopping sequence.

Description

Technical Field

[0001] The present invention relates to wireless transmitters and wireless data networks comprising a plurality of transmitters. In a particular but not exclusive use, the present invention relates to low power measurement nodes for IoT (Internet of Things) applications and IoT wide area networks comprising a plurality of measurement nodes together with gateways and other devices. The present invention utilizes frequency hopping to extend the spectrum of narrowband radio signals. Background Art

[0002] Several competing low-power radio networks are used to provide connectivity in IoT applications. Local area networks like Wi-Fi and Bluetooth have been used successfully in some applications, but they require local infrastructure to connect to the Internet, which is not always available or desirable, and are rarely suitable for mobile applications where sensor nodes may move outside the accessible range of a Wi-Fi or Bluetooth gateway. Low-power wide-area networks (LPWANs) have also been proposed for IoT applications. These networks typically operate in unlicensed radio bands and must have high interference immunity while maintaining regulatory compliance.

[0003] Frequency Hopping Spread Spectrum (FHSS) is a radio transmission method for increasing the bandwidth of a signal. In this method, the transmitter changes the transmission carrier frequency according to a pseudo-random sequence, the "hopping sequence". By knowing the hopping sequence and using a receiver that is synchronized with the transmitter frequency hopping, the message can be decoded and understood. FHSS transmissions are highly resistant to interference and for this reason are advantageously used in unlicensed frequency bands.

[0004] Low-power wide-area networks are using frequency hopping to increase system capacity at low data rates.

[0005] Transmission by radio means is subject to very complex regulations that vary in different countries and regions. These regulations generally require that the transmission energy should be evenly distributed among the available channels and that the occupancy of each individual channel does not exceed a given limit. For example, FCC rules stipulate that in a general time window of 20 seconds, no channel should be occupied for more than 0.4 seconds and that the system will use at least 50 hopping channels spaced at least 25 kHz apart.

[0006] Pseudo-random sequences are at the heart of many frequency hopping systems and can be generated in a variety of ways. It is desirable that a continuous channel be governed by a sequence with a long period and that all possible values ​​in its domain are assumed at once, i.e., permuted. Linear Feedback Shift Registers (LFSRs) are commonly used to generate such sequences. They generate a series of values ​​that is completely determined by the initial state and is necessarily periodic since the domain over which data can be generated is finite. The exact choice of the reverse-action function produces a sequence of maximum length. Such reverse-action functions can be expressed as modulo-2 polynomials and their coefficients can be represented by numbers with as many bits as the number of shift register cells.

[0007] Most often, transmitters of low-rate networks do not have accurate real-time clocks. They spend most of their time in low-power states, where time is only approximately kept, or not kept at all. When they have to transmit data (e.g., upload measurements to a gateway), synchronization exchanges are preferably avoided because they use network capacity and consume power. Ideally, IoT transmitters should be able to wake up from a low-power state, collect a measurement and transmit it without overhead. Receivers cannot rely on the fact that all transmission channels appear in predetermined time slots, nor on cooperation between transmitters.

[0008] Messages exchanged in low-rate networks do not follow a predictable schedule and often include special detection sequences or synchronization words that are used on the receiver side to determine that the message is arriving. Reliable detection imposes a minimum length on the detection sequence. Summary of the invention

[0009] The invention proposes a radio transmitter that transmits data to a corresponding receiver using frequency hopping spread spectrum modulation. The data is organized into packets, each packet being associated with a special hopping sequence. Each packet contains a series of segments separated by frequency hopping. Each segment is transmitted in a channel determined by the hopping sequence. The packet has one or more header segments, which include a detection sequence and announce the hopping sequence in a suitably coded form, and after the header, a payload segment conveying the desired message. Preferably, each packet has a plurality of redundant headers transmitted one after another on different channels, depending on the hopping sequence.

[0010] The present invention proposes a transmitter configured to generate a suitable hopping sequence for each packet by applying a binary XOR mask to a permutation known to the receiver. In this way, the selected hopping sequence can be signaled in a simple and compact form.

[0011] When there are several transmitters in the network, each transmitter is configured to randomly generate a different hopping sequence and use that hopping sequence with little or no overhead. The hopping sequences should have good cross-correlation properties to limit collisions between different sequences. Correlation is defined as the total number of equivalent frequencies when comparing two frequencies and their relative time or frequency shifts.

[0012] In a preferred variant, the permutation is the output of a maximum length linear feedback shift register whose feedback function is a polynomial from a finite set of maximum length polynomials conventionally known to all receivers in the network. The length of the shift register corresponds to the length of the binary mask. The initial state of the shift register may also be defined by convention. The feedback function may also be announced in the header. Since the number of possible choices is reduced, all that is required is an index into an array known to both the transmitter and the receiver pointing to the selected polynomial.

[0013] The channel sequence can be obtained by repeatedly applying the XOR mask to the state of the shift register and then decrementing the result by one. Since the linear feedback shift register can never produce an empty state, the XOR operation will be omitted when the state of the linear register is the same as the binary mask, otherwise the value of the binary mask can never be generated. This sequence produces a series of pseudo-random numbers with a range of [ 0 , L-2 ], where L is the period of the LFSR. Since the feedback function is chosen to produce a period of maximum length, it maintains L = 2 n ,in n is the length of the shift register.

[0014] Another possible implementation consists in inserting zeros at certain positions in the shift register output and then applying an XOR mask.

[0015] The method used in the transmitter disclosed herein generates as many unique hopping sequences as the product of the number of possible masks multiplied by the selected feedback function subset. For practical and regulatory reasons, the number of available channels may be less than the range of hopping sequences generated by the method of the present invention. In this case, the range of permutations can be reduced to a set of available channels by either skipping unavailable channels or replacing them with silence. The latter option is preferred because it retains the desired cross-correlation characteristics of the LFSR.

[0016] The first fragment of a packet includes a detection sequence or sync word, and a header that announces information needed to track and decode the payload, such as hopping sequence, payload length, modulation coding, etc. Given the important and special nature of the information contained in the header or headers, it is desirable that the header is not interrupted by frequency hopping.

[0017] Therefore, the jump in the header may need to be extended, and the transmitter may stay in the channel longer during the header transmission. This requires some attention to the provisions.

[0018] Band occupancy rules may impose a maximum occupancy time on each channel. For example, the FCC stipulates that no frequency should be used for more than 400 ms in any 20 s period. If the length of the header jump is already greater than half that number, then the immediately subsequent packet selects a sequence in which the same channel is repeated in one of the first positions used in the header, which may violate this rule unless the transmitter delays the subsequent packet for at least 20 s.

[0019] The transmitter of the present invention is configured to select a hopping sequence of packets such that the channel of the first hop in the hopping sequence is different from the channel of the same position in the immediately preceding packet. Thus, the transmitter of the present invention can transmit the second packet without waiting for 20 seconds. The way the sequence is generated ensures that this condition is observed simply by selecting a binary mask that has a predetermined offset from the binary mask of the immediately preceding packet. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The present invention will be disclosed with reference to the accompanying drawings, which illustrate:

[0021] Figure 1 : Two packets transmitted by the first variant of the transmitter of the invention.

[0022] Figure 2 : A packet transmitted by the second variant of the transmitter of the present invention. DETAILED DESCRIPTION

[0023] The present invention relies on the definition of hopping sequences with good spectrum usage and capacity characteristics in low-rate low-power wide area networks. The transmitter of the present invention generates radio messages that are organized into packets, each packet being associated with a special hopping sequence. Each packet contains a series of segments separated by frequency hopping; the carrier frequency is changed according to the hopping sequence. The transmitting devices of the present invention may not be synchronized in time, and random hopping sequences are used to limit collisions.

[0024] A packet begins with at least one header fragment, which contains the detection sequence and header information needed to track and decode the payload data. The header information can vary, but in a typical implementation, they contain the payload length and format and the modulation scheme. Preferably, the header announces the hopping sequence that will follow in the series of fragments that will come later.

[0025] Given the important nature of the information encoded in the header, the header is preferably repeated in a redundant manner. For example, each packet may begin with two, three, or four header fragments. Each header fragment contains enough information to allow a receiver to synchronize with the hopping sequence and decode the contents of the payload fragment.

[0026] The header segments have a minimum size determined by the size of the sync word and the amount of information they must convey. This may require a longer duration than the payload segment duration. In a practical implementation, the "normal" segment containing the payload has a length of 102.4 ms, while the header payload has a length of 233.47 ms for the reasons described above. This is more than half of the maximum dwell time allowed in the FCC regulations, and requires special considerations, which will be disclosed later.

[0027] According to aspects of the invention, a technique for generating a hopping sequence relies on an N-bit linear feedback shift register of a base sequence and uses a bitwise XOR operator to derive further sequences. The transmitter selects a polynomial of maximum length N for the feedback function. The number of bits of the shift register and the polynomial are selected by appropriate conventions based on the number of channels that must appear in the hopping sequence, for example according to the following table.

[0028]

[0029] The table is compiled only for certain sizes of channel grids: 10, 22, 28, 30, 35, 47, 60, 62, 86, 99, 185, 198, 390, and 403. These are interesting or valuable cases made by regulations in certain regions. However, the present invention may include other values. The polynomial and initial state are given in decimal representation.

[0030] Depending on the size of the channel grid, the length N of the shift register is 6, 7 or 8 bits. It will be appreciated that the feedback polynomial is generally more than one. Depending on the size of the channel grid, one or several (up to 6) polynomials are used. The transmitter of the present invention is configured to randomly select a polynomial among the available possibilities. However, it is important and preferred that the polynomial is selected to give a maximum periodic sequence.

[0031] During the sequence, the shift register state will assume [ 1 , 2N-1] (due to linearity, a state consisting of N zeros is missing). The initial state of the LFSR is also formulated and is either 6 or 64 depending on the number of channels. The choice of polynomial and initial state is conventional and can be replaced by others. Note that the initial state is given before the initial step of shifting bits that updates the shift register. Therefore, the initial state of 64 for the second column is valid even though its binary representation is a 7-bit number.

[0032] Therefore, the transmitter's LFSR can generate as many [ 1 , 2N-1 ] replacement.

[0033] In order to increase the number of hopping sequences that can be generated, the transmitter has a step of combining the LFSR base sequence with a predetermined N-bit binary mask in a bitwise XOR operation. The binary mask will be indicated by xor_seed. It can assume all values ​​[ 0 , 2N-1 ] and combined with the N-bit state of the shift register at each step. Because the shift register never generates the value 0, the result will never contain the value xor_seed, which is undesirable. Therefore, when the LFSR state is equal to xor_seed, the algorithm skips the XOR operation and then decrements the result by one to produce [0, 2 N -2]. In terms of code, this part of the algorithm can be coded as:

[0034]

[0035] The update of the shift register (process ) are not described in detail and can be implemented in any suitable known manner, for example using a classical Galois implementation.

[0036] These sequences are particularly easy to generate and store. On average, they use all available channels, regardless of whether the sequence is truncated (which is often the case), and have equal or better cross-correlation properties than randomly generated sequences, i.e., fewer repeated collisions between 2 frames on average.

[0037] In this way, the transmitter can generate the jump sequence m × 2 N permutations, where m is the number of possible polynomials for the chosen grid size. To put some numbers on the idea, for a grid of 60 channels, the table gives N =6 and m = 6. In this way, the transmitter can 384 =6 × 26 The hop sequence is selected from a pool of hop sequences and announced in one header or in multiple headers. The polynomial and initial state can be chosen so that all hop sequences are unique.

[0038] This way of combining the LFSR state with the XOR binary mask to multiply the permutation number is preferred, but not exclusive, nor necessary. In a possible variant, the same result can be obtained by appending zero values ​​at predetermined positions in the sequence, followed by XORing the LFSR state with the binary mask.

[0039] Optionally, the transmitter may introduce additional bit operations on the result, such as a bit shuffle operation.

[0040] Keeping the values ​​given above as examples, the generated pseudo-random sequence will have a range [0, 62]. This can be reduced to the desired number of channels, for example 60, by skipping non-existent channels or preferably replacing non-existent channel codes with silence intervals. The latter option is preferred because, as can be shown, it preserves the cross-correlation properties of the original sequence generated by the LFSR, which is advantageous, and reduces the number of random collisions.

[0041] To provide a numerical example, we will assume that there are 60 available channels. A hopping sequence can be constructed by generating a base pseudo-random random sequence of length using the provisions of Table 1, and then altering it by an XOR operation with the mask. Selecting the polynomial "51" between those available and the binary mask 0b011000 (decimal 24), the hopping sequence will be:

[0042]

[0043] The same process, only changing the binary mask to 0b011100 (28) produces:

[0044]

[0045] As already briefly mentioned, the header fragment is longer than the payload fragment and may exceed half of the maximum dwell time allowed by FCC rules in any 20 s period. This may lead to a rule violation because the probability of transmitting two header fragments in the same channel is non-zero if two packets are transmitted one after the other with different hopping sequences selected using the method disclosed above. This can be avoided using a simple but inefficient device that spaces the packets by 20 s or more.

[0046] The present invention proposes a preferred solution, namely to select the hopping sequence of packets in such a way that no channel in the initial subset of consecutive channels of the sequence appears in the same subset of channels of the immediately preceding hopping sequence. In this way, the transmitter can transmit two packets separated by only 10 s, and it is certain that no two header fragments will happen to be in the same radio channel.

[0047] The generation method of the present invention provides a particularly simple way to select a non-repeating skip sequence simply by selecting a binary mask xor_seed that is shifted by a determined amount from the previous binary mask. It can be shown that if the value of xor_seed differs 4. 8. 120. 136 or 184, there will be no repetition in the first three elements of the skip sequence. If they differ 8. 120. 136 or 184, then there will be no repetition in the first four elements of the hopping sequence. This very simple rule ensures that the requirements of FCC Part 15 are met while transmitting one packet every 10 seconds.

[0048] Figure 1 This arrangement is illustrated. A transmitter sends a first message using the hopping sequence of Table 2. This is for an FCC regulated region with 60 channels on a conventional grid. The header fragments H0, H1, H2 are longer than the payload fragments P0-P12. After a delay of 10 seconds, the same transmitter sends a second message using the sequence of Table 3. In this case, since the bit masks differ by 4, and the polynomials are the same, the header packets H0-H2 are inherently generated on a different channel than the previous packet channel. There is no such guarantee for the payload packets, but since these packets are shorter, the rules regarding maximum dwell time are followed.

[0049] A limitation of the above arrangement is that, since the receiver cannot synchronize with the hopping sequence without the important information included in the header (sync word, hopping sequence announcement), it is necessary that this information is transmitted in a single segment without hopping, the length of which cannot be reduced arbitrarily. A solution to this additional problem would be to make the hopping sequence into two nested sequences.

[0050] The inner sequence is a short sequence determined by a simple invariant rule, and its jumps are relative to the starting channel (in other words, they are incremental jumps) so that the receiver can follow it without knowing the full hop sequence. The outer sequence will have a range of a small subset of the available channels, and the inner sequence is chosen to use all available channels when nested with the outer sequence.

[0051] Now we will disclose a way to generate a nested hopping sequence using all available channels, but this is just one example among many examples included in the present invention. First, we select the length of the inner sequence " k This must be a divisor of the number of available channels and determines the maximum number of fragments that can be received without synchronizing to the full hop sequence.

[0052] The internal sequence is an unchanging sequence of numbers , which have the following characteristics: they are k The remainders of the divisions are all different, or equivalently, yes Without loss of generality, choose It's always possible.

[0053] Depend on The external sequence indicated will have length ,in N is the number of available channels, and k is the length of the internal sequence. It can be constructed as disclosed above by first generating a sequence of length The base permutation of and then by the base permutation and the bit mask to change it. In this variant, the scope of the outer sequence is only a small subset of the available channels.

[0054] Then by multiplying the value of the outer sequence by k And the increments of the inner sequence are added successively to construct the full sequence, thus:

[0055]

[0056] The operator "\" represents integer division. It can be proved that the whole sequence is replacement.

[0057] To revise our thinking, we will again assume that there are 60 available channels and that the required length of the inner sequence is k = 3 For example, the internal sequence is chosen to be The outer sequence can be constructed by generating a base random sequence of length 20 and then altering it by an XOR operation with the mask. As an example, the outer sequence can be:

[0058]

[0059] It is obtained by using a 6-bit LFSR, polynomial 51 (0b110011), initial state 6, and XOR mask 24 (0b011000). This is consistent with the internal sequence Nesting produces:

[0060]

[0061] Importantly, these sequences are not completely random, because when the elements are grouped by k=3, each group has of form. However, they use all available channels equally and their correlation properties are sufficient. If a group of three channels is a header and the receiver detects the first fragment (for example, by a synchronization word embedded in it), the channels of the two subsequent fragments can be determined from the first fragment without knowledge of the full hop sequence. The synchronization word can also be split into 2 or 3 fragments, in which case the receiver can combine them before detection due to the deterministic hops between the fragments. Thus, the transmitter can send a long header without increasing the length of the fragment and without exceeding the specified dwell time on each channel.

[0062] Figure 2 This variation is illustrated. Each group of three fragments follows the same pattern, as determined by the internal sequence. The groups are randomly distributed according to the external sequence. The packet has two headers: the first header consists of fragments H0 H1 H2, and the other header consists of H3 H4 H5. For robustness, the second header can be a redundant repetition of the first header, or convey different information as needed.

[0063] This specification discloses specific embodiments of the present invention. However, various modifications and variations may be made without departing from the spirit and scope of the present invention as defined in the claims. Therefore, the specification and drawings should be regarded as illustrative rather than restrictive.

Claims

1. A radio transmission device configured to transmit a spread spectrum radio signal in which a carrier frequency changes in a predetermined set of radio channels according to a hopping sequence, the radio signal being organized into packets, each packet having a header transmitted at a first channel in the hopping sequence - the header including a detection sequence, and payload data encoding a message transmitted at subsequent channels in the hopping sequence, the radio transmission device being configured to start transmitting the header by selecting a binary mask, and to generate the hopping sequence by: generating a pseudo-random permutation of numbers, altering the pseudo-random permutation by bitwise XORing with the binary mask, reducing the range of the altered permutation so that the range is contained in the set of radio channels, encoding the binary mask in the header to announce the hopping sequence.

2. The radio transmission device according to claim 1, in, The header announces the hopping sequence.

3. The radio transmission device according to claim 1, in, The packets have redundant headers which announce the hopping sequence transmitted at different channels. The radio transmitting device according to claim 1 , configured to shuffle bits of a hopping sequence.

5. The radio transmitting device according to claim 1, configured to generate a pseudo-random permutation of numbers having a maximum length LFSR, wherein a feedback function of the LFSR is encoded in the header.

6. The radio transmission device according to claim 1, in, Reducing the range to the set of channels includes skipping values ​​that are not in the set of channels, or inserting silence when a value is not in the set of channels.

7. The radio transmission device according to claim 1, in, The pseudo-random permutation change is the XOR of the LFSR state with the binary mask - provided that the LFSR state is not equal to the binary mask, and then decremented by one.

8. The radio transmission device according to claim 1, in, A variation of the pseudo-random permutation is the addition of zero values ​​at predetermined positions in the sequence, followed by an XOR of the state of the LFSR with a binary mask.

9. The radio transmission device according to claim 1, in, The binary mask is selected in such a way that the channels in the first subset of consecutive channels at the beginning of the hopping sequence are different from the channels that appeared in the same subset of channels in the hopping sequence of the immediately preceding packet.

10. The radio transmission device according to claim 9, in, The header or the redundant header has a residence time at the corresponding channel that is longer than the residence time of the payload data.

11. The radio transmission device according to claim 9, in, The binary mask has a predetermined offset from the binary mask selected for the immediately preceding packet.

12. The radio transmission device according to claim 1, in, The hopping sequence includes two nested sequences, namely an outer sequence and an inner sequence, wherein the outer sequence is a sequence generated by changing the pseudo-random permutation through a bit-by-bit XOR operation with a binary mask and reducing the range of the changed permutation in the radio channel set, and the inner sequence is a sequence of constant predetermined increments.

13. The radio transmission device according to claim 12, in, The outer sequence has a range that is a proper subset of the radio channels, and the inner sequence is selected such that the hopping sequence uses all radio channels.

Images