Adaptive modulation anti-multipath unmanned aerial vehicle swarm communication method and system based on MIMO-OFDM (Multiple Input Multiple Output-Orthogonal Frequency Division Multiplexing)

Abstract

The invention discloses an adaptive modulation anti-multipath unmanned aerial vehicle swarm communication method and system based on MIMO-OFDM, and relates to the technical field of wireless communication, and the method comprises the steps: receiving a down-conversion baseband; synchronization and compensation are carried out; removing a cyclic prefix and carrying out fast Fourier transform; estimating a channel and noise on the training and pilot frequency, and correcting a common phase; performing frequency domain equalization and soft decision, forward error correction and cyclic redundancy check according to the estimation to obtain a control parameter, decoding a payload according to the control parameter, and performing statistics to form a link quality indication; modulating coding, bandwidth and spatial mode, configuring mapping, power and weight, and packaging physical layer protocol data; carrier sensing conflict avoiding sending, confirmation and retry are adopted to generate media access statistics; and the routing selects the next-hop multi-hop forwarding according to the transmission quality, and performs end-to-end statistics and recharge. Through cooperation of hierarchical synchronization and phase correction with cyclic prefix and FFT, crosstalk is suppressed, channel noise scale optimization soft decision is robust, data decoupling is controlled, and end-to-end availability is improved through multi-hop self-healing recharge.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of wireless communication, in particular to a MIMO-OFDM-based adaptive modulation anti-multipath UAV swarm communication method and system. BACKGROUND

[0002] In the application of low-altitude near-far cooperative reconnaissance, emergency communication relay, formation operation, etc., the UAV swarm puts forward the requirements of robust reliability and controllable delay for networking communication under the conditions of strong multipath, non-line-of-sight, fast maneuvering and dense concurrent. The following common challenges exist in such an environment: first, low-altitude near-ground propagation leads to significant delay spread and frequency-selective fading, and traditional single-carrier systems are easily affected by inter-code interference without complex equalization; second, platform maneuvering and local oscillator imperfection introduce carrier frequency offset and phase noise, which easily causes inter-carrier interference (ICI) in orthogonal frequency division multiplexing (OFDM) systems; third, dense access of the swarm triggers strong competition and collision, leading to degradation of effective throughput and delay jitter; fourth, dynamic changes in multi-hop topology make the end-to-end path quality fluctuate over time, further amplifying the uncertainty of upper-layer services.

[0003] In the prior art, OFDM approximates diagonalization of linear convolution through cyclic prefix, and completes channel estimation with pilot, which can resist multipath to a certain extent; MIMO improves reliability or spectral efficiency through diversity, multiplexing or beamforming; CSMA / CA realizes distributed access through physical / virtual carrier sensing and random backoff; and a number of ad hoc routing protocols realize next-hop selection through routing table and link metric. However, in the complex scenario of UAV swarm, the existing solutions generally have the following shortcomings: the suppression link of coarse frequency offset and residual phase noise is often not closed, the common phase error (CPE) is not continuously tracked at the symbol level, which limits the performance of subsequent equalization and soft decision; common solutions only make one-dimensional switching of modulation and coding (MCS), lack of linkage with bandwidth (BW) and MIMO spatial processing mode (diversity / multiplexing / beam); lack of hysteresis and minimum dwell constraints, leading to system instability; if the next-hop selection at the routing layer does not introduce transmission quality parameters and aging mechanisms consistent with the physical / access layer, it is difficult to recover in time when the path quality fluctuates; and if the end-to-end statistics are not backfilled to adaptive control, it is impossible to correctly adjust the throughput bottleneck.

[0004] In summary, the prior art lacks an integrated method that penetrates the three layers of physical-access-routing, forms a closed loop in the full link, and realizes stable, controllable and efficient swarm communication under multipath and maneuvering conditions. To solve the above problems, it is necessary to propose a robust physical layer that uses short / long preamble hierarchical synchronization and two-stage frequency offset compensation, joint noise estimation of training and null subcarriers, and symbol-level CPE tracking; a self-adaptation of MCS, bandwidth and spatial mode three-dimensional linkage driven by a link quality indication set of physical / access / end-to-end three-layer fusion; a cross-layer closed loop based on transmission quality for routing maintenance and multi-hop forwarding and statistical backfilling to realize robust communication in the strong multipath, strong competition and dynamic topology environment of UAV swarm. SUMMARY

[0005] Based on the above-mentioned shortcomings of the prior art, the purpose of the present application is to provide a MIMO-OFDM-based adaptive modulation anti-multipath UAV swarm communication method and system to solve the above technical problems.

[0006] To achieve the above-mentioned purpose, the present application provides the following technical scheme: a MIMO-OFDM-based adaptive modulation anti-multipath UAV swarm communication method, comprising:

[0007] Receiving a complex baseband sample sequence after radio frequency down-conversion, denoted as a first data stream; performing coarse timing estimation on the first data stream to generate a coarse synchronization parameter set and a first compensated data stream; performing fine timing estimation on the first compensated data stream to generate a fine synchronization parameter set and a second compensated data stream;

[0008] According to the fine synchronization parameter set and the system configuration set, the second compensated data stream is removed from the cyclic prefix symbol by symbol and 64-point FFT is performed to obtain a subcarrier frequency domain symbol matrix, denoted as a first frequency domain data; the subcarrier level initial channel estimation matrix and the noise power estimation are calculated according to the first frequency domain data; the common phase error sequence is calculated based on the first frequency domain data and the noise power estimation and phase compensation is performed to obtain the second frequency domain data;

[0009] The subcarrier level initial channel estimation matrix and the noise power estimation are used to implement frequency domain equalization and soft demapping on the second frequency domain data symbol, and the control parameter set is obtained through forward error correction decoding and CRC check; according to the control parameter set, equalization decoding is performed on the data symbol in the second frequency domain data to obtain the payload bit sequence, the link quality parameter is collected, and the link quality indication set is generated;

[0010] According to the link quality indication set and the queue state, a target modulation and coding scheme, a target bandwidth and a target spatial processing mode are selected in the modulation and coding set to form a control triple; according to the control triple, the mapping, power and MIMO weight / codebook of the sender are configured, and the payload bit sequence is packaged as a physical layer protocol data to be sent;

[0011] performing sending of physical layer protocol data to be sent, returning an acknowledgement frame upon successful reception, and retrying according to rules upon failure, generating medium access statistics;

[0012] selecting a next hop according to a routing table and a transmission quality parameter at a self-organizing network routing layer, performing multi-hop forwarding of a payload bit sequence to form an end-to-end transmission, and generating end-to-end statistics to be fed back to a set of link quality indicators together with the medium access statistics.

[0013] The application further provides that a sliding autocorrelation window with a delay of an integer multiple of 16 is set, a delay correlation quantity and a corresponding energy are calculated, a detection metric is constructed, the detection metric is a ratio of a square of a magnitude of the delay correlation quantity to a square of the corresponding energy, platform entry is determined based on a threshold and a minimum holding length, a first timing position is determined according to a maximum criterion of a correlation magnitude or energy within an entry interval, a first frequency offset is obtained according to a ratio of a complex phase of the correlation quantity to the delay, de-rotation compensation is performed on the first data stream, a coarse synchronization parameter set composed of the first timing position and the first frequency offset is outputted, and a first compensated data stream is outputted.

[0014] The first compensated data stream is subjected to a sliding cross-correlation with a local long preamble as a template, a maximum correlation peak value exceeding a threshold is searched for within each of two adjacent intervals with a length of 64, a second timing position is determined according to a first peak position, a second frequency offset is obtained according to a ratio of a phase difference of two peak complex correlation values to the length of 64, de-rotation compensation is performed on the first compensated data stream, a fine synchronization parameter set composed of the second timing position and the second frequency offset is outputted, and a second compensated data stream is outputted.

[0015] The application further provides that the system configuration set is loaded before a task is started and serves as a common priori at both ends of a transceiver, and is composed of a limited parameter subset, including:

[0016] The physical layer fixed parameter subset includes: a transmission system is MIMO-OFDM; a fast Fourier transform point number is 64; a forward error correction coding type is convolution code; a maximum number of transmitting and receiving antennas is not more than 4; a peak-to-average ratio suppression mode is amplitude limiting; a transmitting power control and a channel detection function are enabled;

[0017] The physical layer switchable gear subset includes: a subcarrier spacing value set is {78.125kHz, 156.25kHz, 312.5kHz}; a cyclic prefix length set is {16, 24, 32, 40, 48, 56, 63} and a pair combination thereof; a modulation mode set is {BPSK, QPSK, 16QAM, 64QAM}; and a physical channel bandwidth set is {5MHz, 10MHz, 20MHz, 40MHz}.

[0018] The spatial processing capability subset includes: a MIMO mode set is limited to spatial diversity, beam forming and spatial multiplexing.

[0019] Access layer timing priori subset: medium access control adopts CSMA / CA, sets short interframe interval and time slot time.

[0020] The application is further configured to take the second timing position in the fine synchronization parameter set as a symbol starting reference, strip the cyclic prefix of the second compensated data stream symbol by symbol according to the cyclic prefix length in the system configuration set, and complete the transformation according to the 64-point fast Fourier transform size in the system configuration set to obtain a subcarrier frequency domain symbol matrix covering the preamble, signaling and data symbols, denoted as first frequency domain data, while retaining the symbol delimiting index for subsequent processing;

[0021] According to the training symbol position and its sending reference predefined in the system configuration set, the ratio of the received value to the reference value is linear least square estimated, and time averaging is performed among multiple training symbols to obtain the subcarrier-level initial channel estimation matrix;

[0022] Based on the energy averaging of the empty subcarrier set defined in the system configuration set, or based on the energy estimation between the residual of the received value of the training symbol and the predicted value of the initial channel estimation, one or both of them can be selected or fused to obtain the noise power estimation;

[0023] According to the pilot subcarrier position and pilot symbol given in the system configuration set, and combining the initial channel estimation, a single common phase error is calculated for each orthogonal frequency division multiplexing symbol; when the pilot confidence is insufficient, the inter-symbol smoothing or extrapolation method is used to stabilize the common phase error;

[0024] For all subcarriers of each orthogonal frequency division multiplexing symbol, phase compensation is performed according to the reverse rotation of the common phase error, and a compensated subcarrier frequency domain symbol matrix is output, denoted as second frequency domain data.

[0025] The application is further configured to calculate the subcarrier-level equalization weight according to the subcarrier-level initial channel estimation matrix and the noise power estimation at the control field symbol position according to the linear minimum mean square error criterion to obtain the equalization symbol sequence of the control field; according to the preset control field modulation and mapping rule, the bit log likelihood ratio soft demapping is performed on the equalization symbol, and the deinterleaving and forward error correction decoding are completed; the control parameter set is obtained through the cyclic redundancy check, and the control parameter set at least includes the modulation order, coding rate, interleaving and mapping scheme, spatial stream number and stream mapping, scrambling and frame structure information of the data field;

[0026] At the symbol position of the data field, equalization is performed on the second frequency domain data sub-carrier by sub-carrier according to the space processing mode and modulation coding scheme indicated by the control parameter set; when the space diversity mode, the diversity combining is performed to form a single stream equivalent symbol; when the space multiplexing mode, the multi-stream detection weight is constructed according to the linear minimum mean square error criterion and the stream level symbol estimation is output; when the beam forming mode, the equalization is completed according to the single stream equivalent channel;

[0027] According to the modulation and mapping rule determined by the control parameter set, the bit log likelihood ratio of the equalized symbol is generated, and the forward error correction decoding of the deinterleaving and rate matching is performed, and then the descrambling and cyclic redundancy check are completed, and the payload bit sequence is output; in the multi-stream scenario, the code word assembly is completed according to the code word and stream mapping relationship indicated by the control parameter set;

[0028] In the equalization, soft demapping and decoding process, the physical layer parameters are collected, including: constellation error based on pilot or decision assistance, bit log likelihood ratio amplitude mean or quantile statistics, block error rate or frame error rate, received signal strength or equivalent signal to noise ratio, single hop effective throughput measured by window, and noise power and common phase shift amplitude;

[0029] The physical layer parameter statistics are aggregated into a link quality indicator set according to the field and sub-carrier.

[0030] The application is further provided to determine the target modulation coding scheme, the target bandwidth and the target space processing mode in the modulation and coding set, the bandwidth set and the space processing mode set under the joint constraint of the link quality indicator set and the queue state, wherein the target modulation coding scheme is determined according to the block error rate target before decoding and the equivalent signal to noise ratio threshold, and the hysteresis and minimum residence period are set, the target bandwidth is determined according to the busy time ratio, the retransmission rate or the queue backlog threshold, and the hysteresis is synchronized with the target modulation coding scheme, and the target space processing mode is determined according to the channel rank and the correlation measure, and is constrained by the noise power estimation and the block error rate;

[0031] The modulation order, the coding rate, the interleaving and the bit mapping rule are set according to the target modulation coding scheme; the number of spatial streams, the code word to stream mapping and the diversity combining or multiplexing decoupling strategy are determined according to the target space processing mode; the sub-carrier occupation and the pilot pattern are configured according to the target bandwidth.

[0032] Within the spectrum mask and the maximum power generation limit, the transmit power plan is determined according to the target receiver signal to noise ratio and the error code constraint of the link quality indicator set, and the peak to average ratio suppression is performed on the to-be-sent symbol sequence to reduce the nonlinear distortion.

[0033] When the target spatial processing mode is spatial diversity, loading diversity combining or space-time coding weights; when spatial multiplexing, selecting or updating precoding matrix according to subcarrier level initial channel estimation matrix and allocating power of each stream; when beamforming, selecting or calculating beam weight vector from codebook and generating corresponding index;

[0034] Forward error correction coding and adding cyclic redundancy check to the payload bit sequence according to the target modulation and coding scheme, performing interleaving and bit-to-constellation mapping, inserting pilots and guard subcarriers according to the target bandwidth and target spatial processing mode and completing stream-by-stream symbol assembly;

[0035] Assembling the preamble and training field, signaling or control field and data field to obtain the physical layer protocol data to be sent, and the signaling or control field explicitly carries the target modulation and coding scheme, target bandwidth and target spatial processing mode, codebook or precoding index, frame length and configuration identification.

[0036] The application further provides that physical carrier sensing and virtual carrier sensing are combined to determine the idle state; the virtual carrier sensing updates the network allocation vector timer according to the frame duration field; only when the network allocation vector is zero and the physical carrier sensing determines the idle state and continuously remains idle for no less than the DCF interframe space, the backoff process is entered;

[0037] A backoff count is randomly extracted as the initial value of the countdown within the current contention window; the countdown is decremented at each time slot, and when the channel is busy, the countdown is frozen and the remaining count is preserved, and when the channel is idle again and continuously remains idle for no less than the DCF interframe space, the countdown is resumed; when the backoff count is decremented to zero, the transmission permission is obtained;

[0038] The physical layer protocol data to be sent is immediately sent after the transmission permission is obtained; the receiving end returns an acknowledgement frame after a short interframe space when successfully receiving and completing the verification; the sending end determines a failure event when the acknowledgement frame is not received within a preset acknowledgement timeout;

[0039] The extended interframe space is entered into silence after the failure event occurs, and the extended interframe space timer expires and the channel satisfies the idle determination condition, and then the backoff process can be entered again; when the last failure is not obtaining the transmission right and is not having sent and not being acknowledged, the remaining backoff count before freezing is used; when the last failure is having sent and not being acknowledged, the backoff count is extracted again and the contention window is updated according to the implementation strategy; when the maximum number of retries is reached, the frame transmission is finally determined to fail.

[0040] The application further provides that media access statistics are accumulated within a sliding time window, including at least the number of attempts, the number of successes, the number of retransmissions, the number of acknowledgement timeouts, the average backoff delay, the number of freezes and the length of the extended interframe space silence; the media access statistics are time-stamped and configured with an identification, and are aligned with the access layer statistics entries in the link quality indicator set.

[0041] The application is further configured to update the transmission quality parameters of the neighbors and candidate paths with periodic route advertisement packets, establish and maintain route table entries including destination address, next hop, transmission quality value and invalidation time, and execute elimination on the entries not refreshed beyond the invalidation time;

[0042] On each destination address, the transmission quality values of the candidate next hops are compared to select the working next hop according to the maximum transmission quality criterion;

[0043] The payload bit sequence is hop-by-hop forwarded according to the working next hop shown in the route table; the table lookup and forwarding are repeated at the relay nodes until the destination is reached; when the working next hop is missing or unreachable, path reevaluation is performed and the forwarding is continued with the updated entry;

[0044] The end-to-end time stamps and hop-by-hop results of the sent and arrived packets are recorded, the end-to-end effective throughput, arrival success rate, delay and jitter, path switching frequency and approximate value of expected transmission times are calculated, and the statistical vector and time stamp are output within a sliding time window.

[0045] The end-to-end statistics and media access statistics are aligned and bound to the configuration identifier on the time window, and are assembled together with the physical layer statistics as the end-to-end layer and access layer entries of the link quality indication set.

[0046] The application also provides a MIMO-OFDM-based adaptive modulation anti-multipath UAV swarm communication system for implementing the above-mentioned MIMO-OFDM-based adaptive modulation anti-multipath UAV swarm communication method, which comprises:

[0047] The synchronization processing module receives the complex baseband sample sequence after radio frequency down-conversion, denoted as a first data stream; performs coarse timing estimation on the first data stream to generate a coarse synchronization parameter set and a first compensated data stream; performs fine timing estimation on the first compensated data stream to generate a fine synchronization parameter set and a second compensated data stream;

[0048] The channel phase estimation module removes the cyclic prefix from the second compensated data stream symbol by symbol according to the fine synchronization parameter set and the system configuration set, and performs 64-point FFT to obtain a subcarrier frequency domain symbol matrix, denoted as a first frequency domain data; calculates a subcarrier-level initial channel estimation matrix and a noise power estimation according to the first frequency domain data; calculates a common phase error sequence based on the first frequency domain data and the noise power estimation and performs phase compensation to obtain second frequency domain data;

[0049] Equalization decoding module: implement frequency domain equalization and soft demapping on the second frequency domain data symbol with the initial channel estimation matrix and noise power estimation at subcarrier level, get the control parameter set through forward error correction decoding and CRC check; perform equalization decoding on the data symbol in the second frequency domain data according to the control parameter set to get the payload bit sequence, collect the link quality parameter to generate the link quality indication set;

[0050] Transmit end encapsulation module: select the target modulation and coding scheme, target bandwidth and target spatial processing mode in the modulation and coding set according to the link quality indication set and queue state, and form a control triplet; configure the transmit end mapping, power and MIMO weight / codebook according to the control triplet, and encapsulate the payload bit sequence into the to-be-sent physical layer protocol data;

[0051] Access sending module: send the to-be-sent physical layer protocol data; return an acknowledgement frame when successfully received, and retry according to the rules when failed; generate media access statistics;

[0052] Routing and forwarding module: select the next hop according to the routing table and transmission quality parameter at the ad hoc network routing layer, implement multi-hop forwarding on the payload bit sequence to form an end-to-end transmission; generate end-to-end statistics, and jointly backfill the link quality indication set with the media access statistics.

[0053] The application provides a MIMO-OFDM-based adaptive modulation anti-multipath UAV swarm communication method and system, which comprises the following steps: receiving a complex baseband sample sequence after radio frequency down-conversion, and recording the sequence as a first data stream; performing coarse timing estimation on the first data stream to generate a coarse synchronization parameter set and a first compensated data stream; performing fine timing estimation on the first compensated data stream to generate a fine synchronization parameter set and a second compensated data stream; removing the cyclic prefix of the second compensated data stream symbol by symbol according to the fine synchronization parameter set and a system configuration set, and performing 64-point FFT to obtain a subcarrier frequency domain symbol matrix, which is recorded as a first frequency domain data; calculating a subcarrier-level initial channel estimation matrix and a noise power estimation according to the first frequency domain data; calculating a common phase error sequence based on the first frequency domain data and the noise power estimation and performing phase compensation to obtain second frequency domain data; performing frequency domain equalization and soft demapping on the second frequency domain data symbols based on the subcarrier-level initial channel estimation matrix and the noise power estimation, and performing forward error correction decoding and CRC check to obtain a control parameter set; performing equalization decoding on the data symbols in the second frequency domain data according to the control parameter set to obtain a payload bit sequence, collecting link quality parameters to generate a link quality indication set; selecting a target modulation and coding scheme, a target bandwidth and a target spatial processing mode from a modulation and coding set according to the link quality indication set and a queue state to form a control triplet; configuring a sending mapping, power and MIMO weight / codebook according to the control triplet, encapsulating the payload bit sequence into a to-be-sent physical layer protocol data; sending the to-be-sent physical layer protocol data; returning an acknowledgement frame when the sending is successful, and retrying according to a rule when the sending fails; generating medium access statistics; selecting a next hop according to a routing table and a transmission quality parameter at a self-organizing network routing layer, and performing multi-hop forwarding on the payload bit sequence to form an end-to-end transmission; generating end-to-end statistics, and jointly feeding the statistics back to the link quality indication set, which has the beneficial effects of:

[0054] 1. Anti-multipath and frequency offset robustness improvement: The hierarchical synchronization link adopting short-preamble autocorrelation coarse synchronization, long-preamble cross-correlation fine synchronization and symbol-level common phase error (CPE) compensation, and the symbol delimiting and frequency domain processing by cyclic prefix and 64-point FFT can effectively suppress inter-carrier interference (ICI) in the presence of significant delay spread and frequency offset, improve the condition number and convergence stability of the equalization matrix, and reduce the sensitivity to timing deviation;

[0055] 2. Robust matching of channel and noise scale: The subcarrier-level initial channel estimation is obtained by least squares combined with time averaging and frequency domain smoothing on the training symbols, and the consistent soft metric scale is formed by the two-way noise power estimation of the energy of the empty subcarriers and the energy of the training residual, so that the log-likelihood ratio of the soft decision is more matched with the decoder metric, and the block error rate performance in the low signal-to-noise ratio and frequency selective scene is improved;

[0056] 3. Reliable decoding with control and data plane decoupling: control field pre-equalization and soft demapping, get control parameter set after FEC+CRC, then drive data plane mode-aware equalization and decoding, avoid error propagation caused by parameter mismatch; stable soft information input under three types of spatial processing: diversity, multiplexing and beamforming, improve overall decoding reliability;

[0057] 4. Multi-hop self-healing and end-to-end availability improvement: the routing layer maintains candidate paths based on transmission quality parameters and selects the next hop accordingly, quickly falls back and re-evaluates when the path degrades or fails; end-to-end success rate, delay / jitter and effective throughput are fed back to the link quality indicator set, so that the adaptive strategy can locate the bottleneck and self-heal under the priority of "routing first, then physical", improve overall reachability and continuity.

[0058] The above description is only a summary of the technical solutions of the present application, in order to more clearly understand the technical means of the present application, which can be implemented according to the content of the specification, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the following will describe the specific embodiments of the present application. BRIEF DESCRIPTION OF DRAWINGS

[0059] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings needed in the embodiment description. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor. In the drawings:

[0060] Figure 1 The flow chart of the adaptive modulation anti-multipath UAV swarm communication method based on MIMO-OFDM for an exemplary embodiment of the present application is shown;

[0061] Figure 2 The receiving end signaling symbol equalization and decoding processing schematic diagram for an exemplary embodiment of the present application is shown;

[0062] Figure 3 The receiving end data symbol equalization and decoding processing schematic diagram for an exemplary embodiment of the present application is shown;

[0063] Figure 4 The mesh network unicast scenario diagram for an exemplary embodiment of the present application is shown;

[0064] Figure 5 The mesh network broadcast scenario diagram for an exemplary embodiment of the present application is shown;

[0065] Figure 6 The multi-hop unicast scenario diagram for an exemplary embodiment of the present application is shown;

[0066] Figure 7 A multi-hop broadcast scenario diagram is shown for an exemplary embodiment of the present application. DETAILED DESCRIPTION

[0067] Other advantages and novel features of the present application will be readily appreciated as the same becomes better understood by reference to the following detailed description when considered in connection with the accompanying drawings. It should be noted that the Figures provided herein are by way of illustration only and should not be construed in the context of limiting the scope of the present application.

[0068] It should be noted that the drawings provided herein are merely schematic and that actual implementation of the application can differ from the schematic illustrations as described herein, depending upon specific needs and preferences of a particular implementation. In addition, the drawings provided herein are intended merely to illustrate the general principles of the application and should not be construed in a limiting sense.

[0069] In the following description, numerous specific details are discussed in order to provide a thorough understanding of embodiments of the present application. However, it will be apparent to one of ordinary skill in the art that embodiments of the present application can be practiced without these specific details. In other instances, well-known structures and devices are not described in exhaustive detail in order to avoid obscuring embodiments of the present application.

[0070] Embodiment One:

[0071] An adaptive modulation anti-multipath UAV swarm communication method based on MIMO-OFDM, as shown in Figure 1 includes:

[0072] Receiving a complex baseband sample sequence after radio frequency down-conversion, denoted as a first data stream; performing coarse timing estimation on the first data stream to generate a coarse synchronization parameter set and a first compensated data stream; performing fine timing estimation on the first compensated data stream to generate a fine synchronization parameter set and a second compensated data stream;

[0073] According to the fine synchronization parameter set and the system configuration set, the cyclic prefix of the second compensated data stream is removed symbol by symbol and 64-point FFT is performed to obtain a subcarrier frequency domain symbol matrix, denoted as a first frequency domain data; the subcarrier level initial channel estimation matrix and the noise power estimation are calculated according to the first frequency domain data; the common phase error sequence is calculated based on the first frequency domain data and the noise power estimation and phase compensation is performed to obtain second frequency domain data;

[0074] The initial channel estimation matrix and noise power estimation at subcarrier level are used to implement frequency domain equalization and soft demapping on the second frequency domain data symbol, and control parameter set is obtained through forward error correction decoding and CRC check; the data symbol in the second frequency domain data is equalized and decoded according to the control parameter set to obtain a payload bit sequence, link quality parameters are collected, and a link quality indication set is generated;

[0075] A target modulation and coding scheme, a target bandwidth and a target spatial processing mode are selected from the modulation and coding set according to the link quality indication set and the queue state, and a control triplet is formed; the payload bit sequence is encapsulated into a to-be-sent physical layer protocol data according to the configuration of the source mapping, power and MIMO weight / codebook of the control triplet;

[0076] The to-be-sent physical layer protocol data is sent, an acknowledgement frame is returned when the reception is successful, and a rule is retried when the reception fails; and medium access statistics are generated;

[0077] The next hop is selected according to the routing table and transmission quality parameters at the ad hoc network routing layer, and the payload bit sequence is multi-hop forwarded to form an end-to-end transmission; end-to-end statistics are generated, and are fed back to the link quality indication set together with the medium access statistics.

[0078] The application is further configured to set a sliding autocorrelation window with a delay of an integer multiple of 16, calculate a delay correlation quantity and a corresponding energy, construct a detection metric, which is the ratio of the amplitude square of the delay correlation quantity and the square of the corresponding energy, make a platform entry determination based on a threshold and a minimum holding length, and determine a first timing position in the entry interval according to the maximum criterion of the correlation amplitude or the energy; obtain a first frequency offset according to the ratio of the complex phase of the correlation quantity and the delay, perform de-rotation compensation on the first data stream, and output a coarse synchronization parameter set composed of the first timing position and the first frequency offset and the first compensated data stream; specifically, after the receiving antenna is down-converted by radio frequency and analog-to-digital converted, a complex baseband sample sequence is obtained, which is denoted as the first data stream; an integer multiple delay window with 16 sample points as a unit is set, and is sequentially slid on the first data stream; at each sliding position, the data segment of the current window and the data segment corresponding to the delay position are extracted, the two are multiplied point by point in complex number and accumulated to obtain the delay correlation quantity at the position; the data segment of the current window is multiplied point by point in complex conjugate and squared and accumulated to obtain the energy value of the segment; the energy value of the segment corresponding to the delay position is also obtained by the same method, and the product of the two is taken as the corresponding energy; the amplitude square of the delay correlation quantity is taken as the numerator, and the corresponding energy is taken as the denominator to construct the ratio, which is taken as the detection metric, reflecting the matching degree of the correlation strength and the signal energy; a threshold and a minimum holding length are set, when the detection metric continuously exceeds the threshold and lasts for a certain length, it is considered that the signal is in the effective interval; in the effective interval, the position with the maximum amplitude of the delay correlation quantity or the corresponding energy is selected as the first timing position; the first frequency offset is calculated according to the change of the complex phase of the delay correlation quantity at the first timing position; phase rotation compensation is performed on the first data stream to generate the first compensated data stream, and the coarse synchronization parameter set composed of the first timing position and the first frequency offset is outputted;

[0079] The first compensated data stream is performed sliding cross-correlation with the local long preamble as a template, the maximum correlation peak value exceeding a threshold is searched in two adjacent intervals each with a length of 64, the second timing position is determined according to the first peak position; the second frequency offset is obtained according to the phase difference of the two peak cross-correlation values and the ratio of the length of 64, and the first compensated data stream is performed de-rotation compensation, and the fine synchronization parameter set composed of the second timing position and the second frequency offset and the second compensated data stream are output; specifically, the long preamble signal stored locally is used as the cross-correlation template; on the first compensated data stream, sampling segments equal in length to the template are selected in turn, multiplied point by point and accumulated to obtain a cross-correlation value sequence; in two adjacent intervals each with a length of 64, the maximum correlation peak value exceeding a preset threshold is searched respectively; the first peak position is taken as the second timing position for accurately positioning the symbol starting point; the second frequency offset is calculated according to the 64-point interval ratio between the two peaks by using the phase difference of the two peak cross-correlation values; the first compensated data stream is again performed phase rotation compensation to obtain the second compensated data stream, and the fine synchronization parameter set composed of the second timing position and the second frequency offset is output.

[0080] The application is further provided that the system configuration set is loaded before the task starts and serves as the common priori of the two ends of the transceiver, and is composed of a limited parameter subset, including:

[0081] The physical layer fixed parameter subset: the transmission system is MIMO-OFDM; the fast Fourier transform point number is 64; the forward error correction coding type is convolution code; the maximum number of transmitting and receiving antennas is not more than 4; the peak-to-average ratio suppression mode is limiting; the transmitting power control and channel detection functions are enabled;

[0082] The physical layer switchable gear subset: the subcarrier spacing value set is {78.125kHz, 156.25kHz, 312.5kHz}; the cyclic prefix length set is {16, 24, 32, 40, 48, 56, 63} and its paired combination; the modulation mode set is {BPSK, QPSK, 16QAM, 64QAM}; the physical channel bandwidth set is {5MHz, 10MHz, 20MHz, 40MHz};

[0083] The spatial processing capability subset: the MIMO mode set is limited to spatial diversity, beam forming and spatial multiplexing;

[0084] The access layer timing priori subset: the medium access control adopts CSMA / CA, and the short interframe space and time slot time are set.

[0085] Specifically, CSMA / CA stands for Carrier Sense Multiple Access with Collision Avoidance. When the carrier sense process finds the channel idle, the station can start the channel contention process to compete for the channel.

[0086] Because the carrier sense of the physical layer is not always reliable in a complex mobile environment, the carrier sense mechanism of CSMA / CA includes both the physical layer carrier sense and the MAC layer virtual carrier sense, which are in a cooperative relationship.

[0087] The physical layer carrier sense technology contains two objectives: carrier sense and clear channel assessment (CS / CCA). The function of CS is to detect the start of a valid network signal on the channel; CCA is to determine whether there is an activity on the channel at present, and whether the strength of the activity is enough to interfere with the correct reception of the valid signal. The physical layer, in a non-transmitting or receiving state, can start the CS / CCA mechanism to assess the busy or idle state of the channel by monitoring the physical signal on the channel. The channel assessment method is related to the specific physical layer technology. For example, if the channel signal energy is detected to exceed a specified value within a period of time, it can be determined that the channel is in a busy state.

[0088] The MAC layer virtual carrier sense is not to sense the carrier by judging the signal on the medium, but to achieve a kind of virtual carrier detection from the relevant information carried in the MAC frame. Each MAC frame must carry the duration field, the content of which is the duration of the next MAC frame. After receiving a MAC frame, the station must first take out the duration field in the frame to complete the virtual carrier sense after judging that the destination address of the frame is not the station, and then discard the frame.

[0089] (1) Interframe Space

[0090] Interframe space refers to the length of time between frames in time. The station should select different lengths of interframe space for carrier sensing to determine whether the channel is idle according to the operation mode specified in the CSMA / CA protocol under different circumstances. The size of the interframe space is independent of the data rate of the channel, and it varies according to the different physical layer transmission technologies. We will use the following types of interframe space:

[0091] SIFS: Short Interframe Space;

[0092] Slot Time: A time slice, in CSMA / CA, before the node competes for access to the channel, a corresponding random backoff process is required, in which the backoff process is composed of many time slots.

[0093] DIFS (DCF Interframe Space): DIFS = SIFS + 2 x Slot Time;

[0094] EIFS (Extended Interframe Space): EIFS = SIFS + DIFS + ACK_Duration.

[0095] SIFS is the time interval between two consecutive frames in a frame exchange sequence. After receiving a MAC frame from the opposite station, the station will wait for SIFS time for the channel to be idle before responding with a MAC frame. SIFS is the shortest interval among all the interframe spaces, and its purpose is to give the highest priority to the station in the frame exchange sequence to access the channel, and to ensure that the frame exchange sequence will not be interrupted or disturbed.

[0096] DIFS is the time interval that a station will wait for the channel to be idle before it will contend for the channel. When the carrier sense tells the station that the channel has changed from busy to idle, the station will continue to sense the channel for DIFS time. Only after the channel has been idle for DIFS time, will the station start the random backoff procedure. When a station receives a corrupted MAC frame (including the case that the MAC frame is collided), it will use EIFS instead of DIFS as the time interval to defer from accessing the channel after the channel has changed from busy to idle. That is, the station will continue to sense the channel for EIFS time. Only after the channel has been idle for EIFS time, will the station start the random backoff procedure. If the station later receives a correct MAC frame, it will use DIFS again for the silence period before contending for the channel.

[0097] where SIFS and Slot time are defined as follows:

[0098] aSIFSTime = aRXRFDelay (Radio Frequency Delay) + aRXPLCPDelay (Physical Layer Header Reception Delay) + aMACProcessingDelay (MAC Layer Processing Delay) + aRxTxTurnaroundTime (Transmit / Receive Antenna Turnaround Time);

[0099] aSlotTime = aCCATime (CCA Time) + aRxTxTurnaroundTime (Transmit / Receive Antenna Turnaround Time) + aAirPropagationTime (Air Propagation Delay) + aMACProcessingDelay (MAC Layer Processing Delay).

[0100] (2) Protocol working principle

[0101] The working principle of the CSMA / CA protocol is shown as follows:

[0102] First, if the node has data, it needs to compete for sending, the node first listens to the channel state, if the channel is idle within DIFS time, then backoff for n slotTime, if the channel is always idle during the random backoff process, the node competes for the channel and sends data; if the channel is found to be busy during the backoff process, the random backoff process is suspended, and the current random backoff value is saved;

[0103] If the receiving end correctly receives the data of the sending end, waits for SIFS time, and feeds back an ACK confirmation frame to the sending end, when the sending end correctly receives the ACK confirmation frame, this time of transmission is completed;

[0104] After this time of transmission is completed, the node needs to wait for DIFS time again, and then restart the random backoff process, if the node has just sent data, then a random number n needs to be selected from the contention window again at the beginning of the random process, and n slotTime backoff is restarted. If the node has just not sent data, then the random backoff value saved last time is directly continued. The purpose of such design is to ensure the fairness of network transmission.

[0105] The application is further provided as the second timing position in the fine synchronization parameter set as a symbol starting reference, the second compensated data stream is stripped of the cyclic prefix symbol by symbol according to the cyclic prefix length in the system configuration set, and the transformation is completed according to the sixty-four-point fast Fourier transform size in the system configuration set, to obtain a subcarrier frequency domain symbol matrix covering the preamble, signaling and data symbols, recorded as first frequency domain data, while the symbol delimiting index is reserved for subsequent processing; specifically, the second timing position is taken as the starting point of the first symbol. The starting point of each symbol is sequentially added with the interval of "cyclic prefix length + sixty-four samples" based on the starting point of the previous symbol; at each symbol starting point, the cyclic prefix length consistent with the configuration is discarded, and only the subsequent sixty-four valid samples are reserved; the sixty-four-point transformation is performed on the above sixty-four samples to obtain the complex frequency domain value of the symbol on all subcarriers; the above steps are repeated for all symbols, and the frequency domain vectors of the symbols are spliced in time sequence to form the first frequency domain data; meanwhile, the delimiting index of each symbol is saved as a time reference for subsequent processing;

[0106] According to the training symbol position predefined by the system configuration set and the sending reference, the ratio of the received value to the reference value is linear least square estimated by subcarrier, and time average is performed among multiple training symbols to obtain the subcarrier level initial channel estimation matrix; specifically, according to the training symbol serial number given in the configuration and the sending reference (the known value expected to be sent on each subcarrier), the corresponding frequency domain received value is taken out from the first frequency domain data; for each occupied subcarrier, the received value is compared with the corresponding sending reference one by one to obtain the complex channel coefficient initial value of the subcarrier in the "proportion relationship of the received value to the reference"; if there are multiple training symbols, the multiple estimates of the same subcarrier are averaged in time to reduce noise fluctuation; slight smoothing (such as moving average or low-order fitting) is performed within the subcarrier neighborhood to weaken isolated outliers to obtain the subcarrier level initial channel estimation matrix;

[0107] Based on the energy average of the empty subcarrier set defined by the system configuration set, or the energy estimation between the residual of the received value of the training symbol and the predicted value of the initial channel estimation, one of the two or the fusion value is obtained to obtain the noise power estimation; specifically, the empty subcarrier energy method: on the configuration specified empty subcarrier set (the frequency point not carrying valid information), the frequency energy at the training time is collected and averaged to obtain the noise baseline. This method is simple and direct and is not affected by signal modulation; the training residual method: taking the initial channel estimation and the sending reference as the predicted value, the residual is obtained by subtracting the training received value by subcarrier; the residual energy is averaged at the training time to obtain the noise estimation. This method can reflect the actual equivalent noise and modeling error. If the difference is not large, the weighted average can be taken; if the difference is obvious, the more conservative (larger) estimation is preferred to ensure the safety margin of equalization and soft decision; the estimation result is smoothed in time to avoid short-term fluctuations caused by individual training symbols;

[0108] According to the pilot subcarrier position and pilot symbol given by the system configuration set, and in combination with the initial channel estimation, a single common phase error amount is calculated for each orthogonal frequency division multiplexing symbol; when the pilot confidence is insufficient, the inter-symbol smoothing or extrapolation method is used to stabilize the common phase error amount; specifically, on each data or signaling symbol, the frequency domain received value at the pilot subcarrier position is extracted according to the configuration; the pilot received value is aligned with the “initial channel estimation × pilot transmission reference” to obtain the representative value of the “overall phase offset amount” of the symbol; a plurality of pilots can be weighted and aggregated to improve stability; the phase estimation reliability of the symbol is evaluated according to the pilot received strength and the “coherence index” after aggregation; when the pilot confidence is insufficient, the smoothing (for example, the previous symbol and the current symbol are averaged according to the weight) or extrapolation (the phase trend of the previous symbol is prolonged along the time) of the adjacent symbol is used to stabilize the common phase error amount; if necessary, “decision assistance” (a small amount of decision results of the data subcarriers are used to participate in the aggregation at high signal-to-noise ratio) is enabled to reduce the instability caused by sparse pilots; a “common phase error sequence” is obtained for the whole frame, and each symbol corresponds to a phase compensation amount;

[0109] The phase compensation of all subcarriers of each orthogonal frequency division multiplexing symbol is performed according to the reverse rotation of the common phase error, and the compensated subcarrier frequency domain symbol matrix is output, denoted as the second frequency domain data; specifically, for each symbol, the frequency domain values of all subcarriers of the symbol are inversely rotated according to the common phase error corresponding to the symbol to realize one-time phase correction; the phase compensation amount is limited or slowly varied to prevent unreasonable jumps between adjacent symbols; the compensated frequency domain data is assembled in time sequence to obtain the second frequency domain data.

[0110] The application is further provided as, Figure 2As shown, at the symbol position of the control field, the subcarrier-level equalization weights are calculated according to the linear minimum mean square error criterion based on the subcarrier-level initial channel estimation matrix and noise power estimation, thus obtaining the equalization symbol sequence of the control field; the equalization symbols are soft-mapped using bit-log likelihood ratio according to the preset control field modulation and mapping rules, and deinterleaving and forward error correction decoding are completed; the control parameter set is obtained through cyclic redundancy check, and the control parameter set includes at least the modulation order, coding rate, interleaving and mapping scheme, spatial stream number and stream mapping, scrambling code and frame structure information of the data field; specifically, the receive vector and corresponding channel estimation and noise power of the control field are read according to the subcarrier; based on the linear minimum mean square error criterion, combined with the channel energy and noise scale, A set of complex weights is calculated for each subcarrier; the received data from multiple antennas / multi-branch systems are linearly synthesized into an equalized symbol sequence using these weights; the equivalent noise intensity for each subcarrier is also provided for subsequent soft-decision calibration; the equalized symbols are converted into bit-by-bit log-likelihood information according to the fixed modulation and mapping rules of the control field; the equivalent noise intensity obtained in the previous step is used for calibration; deinterleaving and channel decoding are performed to obtain the control bit stream; cyclic redundancy check is performed on the control bits; if the check passes, a control parameter set is generated; if it fails, it is processed according to a preset strategy (reset threshold, downgrade weights, request retransmission) and a failure flag is given; the modulation order, coding rate, interleaving / mapping scheme, spatial stream number and mapping, scrambling code and frame structure are written into the control parameter set;

[0111] like Figure 3 As shown, at the symbol position in the data field, equalization is performed on the second frequency domain data subcarrier by subcarrier according to the spatial processing mode and modulation and coding scheme indicated by the control parameter set. When it is spatial diversity mode, aggregation is performed to form a single-stream equivalent symbol. When it is spatial multiplexing mode, multi-stream detection weights are constructed according to the linear minimum mean square error criterion and stream-level symbol estimates are output. When it is beamforming mode, equalization is completed according to the single-stream equivalent channel. Specifically, in spatial diversity mode: each receiving branch is weighted and combined according to the channel conjugate to obtain a single-stream equivalent symbol sequence; in spatial multiplexing mode: a linear minimum mean square error detector is constructed on each subcarrier to output multi-stream symbol estimates; residual interference suppression and per-stream noise recalibration are performed when necessary; in beamforming mode: equalization is performed according to the single-stream equivalent channel, and symbol restoration is completed in combination with precoding weights.

[0112] The modulation and mapping rule determined according to the control parameter set is used to generate bit log-likelihood ratio of the equalized symbol, and forward error correction decoding is performed on the equalized symbol after deinterleaving and rate matching, and then descrambling and cyclic redundancy check are completed, and a payload bit sequence is output; in a multi-stream scenario, code word assembly is completed according to the code word and stream mapping relationship indicated by the control parameter set; specifically, the equalized symbol is converted into bit log-likelihood information according to the modulation and mapping rule indicated by the control parameter set; deinterleaving and code word assembly are performed according to the interleaving / grouping scheme of the control parameter set; in a multi-stream scenario, the order of each code word is restored according to the “code word-stream” mapping relationship; the bit log-likelihood information is input into a channel decoder (the coding rate is given by the control parameter set) of a data plane, and decoded bits are output; descrambling and cyclic redundancy check are completed according to the control parameter set; if the check is passed, a payload bit sequence is output, and if the check fails, a block error is recorded and an upper-layer retransmission mechanism is triggered;

[0113] In the equalization, soft demapping and decoding process, the physical layer parameters are collected, including: constellation error based on pilot or decision assistance, bit log-likelihood ratio amplitude mean or quantile statistics, block error rate or frame error rate, received signal strength or equivalent signal-to-noise ratio, single-hop effective throughput measured by window, and noise power and common phase shift amplitude; specifically, the constellation error: for pilot and (at high signal-to-noise ratio) decision assistance data, the deviation amplitude of the symbol relative to the reference position is calculated and averaged or quantified to obtain the constellation error index; for describing the residual distortion after equalization. Bit log-likelihood strength: the absolute value of the bit log-likelihood in each code word or time window is averaged or quantified to reflect the confidence of the soft information. Block error rate or frame error rate: based on the pass / fail criterion of the cyclic redundancy check, the block-level or frame-level error rate is calculated in a sliding window. Received signal strength or equivalent signal-to-noise ratio: the received amplitude before equalization and the known noise power are used to estimate the received strength; or the equivalent signal-to-noise ratio is inferred from the equivalent noise strength output by the equalization. Single-hop effective throughput: the number of payload bits that pass the cyclic redundancy check is accumulated in a fixed time window, and the effective throughput is obtained by dividing the window length. Noise power and common phase shift amplitude: the noise power estimate and the amplitude statistics of the symbol-level common phase error in the current time window are recorded to reflect the phase noise and residual frequency offset status;

[0114] The physical layer parameter statistics are aggregated into a link quality indicator set according to the field and subcarrier.

[0115] The application is further configured to determine target modulation and coding scheme, target bandwidth and target spatial processing mode in modulation and coding set, bandwidth set and spatial processing mode set respectively under joint constraint of link quality indicator set and queue state, wherein the target modulation and coding scheme takes block error rate target before decoding and equivalent signal to noise ratio threshold as criteria and sets hysteresis and minimum dwell period, the target bandwidth takes busy hour ratio, retransmission rate or queue backlog threshold as criteria and synchronizes hysteresis with the target modulation and coding scheme, and the target spatial processing mode takes channel rank and correlation measure as criteria and is constrained by noise power estimation and block error rate; specifically, block error rate before decoding and equivalent signal to noise ratio are extracted from the link quality indicator set, and constellation error and common phase shift are combined to judge whether the physical layer is degraded; in the modulation and coding set given by the system configuration set, a number of feasible levels are obtained according to equivalent signal to noise ratio level and target block error rate threshold; if there is only a small fluctuation compared with the previous version, or the minimum dwell period has not arrived, the original level is kept unchanged; when the block error rate in the continuous window exceeds the threshold and the physical degradation is established, the level is stepped down; when the multi-window is stable and the queue backlog is obvious, the level is stepped up; whether the access is congested and the service is accumulated is judged according to busy hour ratio, retransmission rate and queue length (or arrival rate), and the principle of preferentially condensing bandwidth (such as reducing from larger bandwidth to medium or smaller) and keeping modulation and coding from being degraded is maintained when congestion dominates and physical quality is normal; when physical quality and access are stable and queue continues to accumulate, bandwidth is gradually increased to release throughput, and modulation and coding scheme shares the same hysteresis and dwell framework; channel rank and correlation are calculated based on recent subcarrier level initial channel estimation matrix, and available spatial dimension is determined by combining noise power and block error rate. Spatial diversity priority: when equivalent signal to noise ratio is low, correlation is high or common phase shift fluctuation is large, select diversity to improve reliability; beamforming priority: when there is obvious main direction or codebook weight adaptation degree is high, select beamforming to obtain array gain; spatial multiplexing priority: when channel rank is sufficient, correlation is low and block error rate and phase are stable, turn on multiplexing to improve peak rate and limit the maximum number of streams to be no more than the upper limit of antenna and rank. Any mode only takes effect under the condition allowed by the system configuration set; if the block error rate deteriorates or the phase shift increases in the continuous window after the change, automatically revert to the last stable mode;

[0116] According to the target modulation coding scheme, the modulation order, the coding rate, the interleaving and the bit mapping rule are set; according to the target spatial processing mode, the number of spatial streams, the code word to stream mapping and the diversity combining or multiplexing decoupling strategy are determined; according to the target bandwidth, the subcarrier occupation and the pilot pattern are configured; specifically, the target modulation coding scheme, the target bandwidth and the target spatial processing mode are combined to form a control triplet, and the timestamp, the configuration identifier and the minimum residence timer are written; the change in a single period shall not exceed the preset maximum step (for example, only adjacent step switching is allowed); otherwise, it is truncated; according to the target modulation coding scheme, the modulation order, the coding rate, the interleaving scheme and the bit to constellation mapping rule are set, and the corresponding decoder parameters are prepared. Spatial mapping configuration: diversity mode: load diversity combining or space-time coding weight to form single stream equivalent mapping; multiplexing mode: determine the number of streams, the assembly relationship of code word to stream and the decoupling strategy baseline of the receiving end; beamforming mode: select or update the weight vector according to the available codebook or weight search strategy, and generate the index; according to the target bandwidth, the subcarrier occupation table and the pilot distribution are selected, and the protection subcarrier and the center alignment rule are kept unchanged;

[0117] Within the spectrum mask and the maximum power generation limit, the transmit power plan is determined according to the target receiving end signal-to-noise ratio and the error code constraint of the link quality indicator set, and the peak-to-average ratio suppression is performed on the to-be-sent symbol sequence to reduce nonlinear distortion; when the target spatial processing mode is spatial diversity, the diversity combining or space-time coding weight is loaded; for spatial multiplexing, the precoding matrix is selected or updated according to the subcarrier level initial channel estimation matrix and the power of each stream is allocated; for beamforming, the beam weight vector is selected or calculated from the codebook and the corresponding index is generated; specifically, according to the target receiving end signal-to-noise ratio and the current error code constraint, the spectrum mask and the maximum power generation limit are combined to calculate the transmit power plan. Inter-stream power allocation: diversity or beamforming: default full-power single stream; multiplexing: default equal power allocation, and if necessary, slightly tilt according to the channel conditions (without breaking the regulations and configuration boundaries). Peak-to-average ratio suppression: amplitude limiting or peak clipping is performed on the symbol sequence to be sent to control the instantaneous peak value and avoid distortion and adjacent channel leakage caused by power amplifier nonlinearity;

[0118] The payload bit sequence is forward error correction coded according to the target modulation coding scheme and cyclic redundancy check is added, interleaving and bit to constellation mapping are performed, pilots and guard subcarriers are inserted according to the target bandwidth and target spatial processing mode, and stream-by-stream symbol assembly is completed; specifically, the payload bit sequence is forward error correction coded according to the target coding rate, cyclic redundancy check and necessary tail bits are added; the bit order is scrambled according to the interleaving rule, and the bits are mapped to constellation points according to the modulation mapping rule; in the multiplexing scenario, the symbols are assembled to each spatial stream according to the "code word-stream" relationship; pilots and guard subcarriers are inserted according to the target bandwidth and pilot pattern to ensure the subsequent common phase error tracking and spectrum shaping requirements;

[0119] Assemble the preamble and training field, signaling or control field and data field to obtain the physical layer protocol data to be sent, the signaling or control field explicitly carries the target modulation coding scheme, target bandwidth, target spatial processing mode, codebook or precoding index, frame length and configuration identification; specifically, the preamble and training field, signaling or control field and data field are assembled in turn to form the physical layer protocol data to be sent; the signaling or control field explicitly carries the target modulation coding scheme, target bandwidth, target spatial processing mode, codebook or precoding index, frame length and configuration identification; the receiving end performs configuration consistency check accordingly; when inconsistent, the new triple is refused to be used and re-announcement is requested.

[0120] The application is further provided to jointly determine the idle state by physical carrier sensing and virtual carrier sensing; the virtual carrier sensing updates the network allocation vector timer according to the frame duration field; only when the network allocation vector is zero and the physical carrier sensing determines the idle state and continuously remains for no less than the DCF interframe space, the backoff process is entered; specifically, when any MAC frame is received, the "duration" field is read and the local NAV timer is updated; during the NAV timer is not zero, the channel is determined as busy. When idle, the received energy or preamble characteristics are detected, and if the threshold is exceeded, it is recorded as busy; if the threshold is below and continuously satisfies the idle criterion, it is recorded as idle. Only when the NAV timer = 0 and the CCA state = idle and continuously remains for no less than the DIFS, the backoff process can be entered; otherwise, the waiting is maintained and the busy time is accumulated;

[0121] The backoff count is randomly extracted as the initial value of the countdown in the current contention window; the count is decremented at each time slot, and when the channel is busy, the countdown is frozen and the remaining count is preserved, and when the channel is idle again and continuously remains for no less than the DCF interframe space, the countdown is resumed; the transmission permission is obtained when the backoff count is decremented to zero; specifically, an integer is randomly extracted as the initial value of the backoff count in the current CW range; the CW initial value is the minimum window; every time a time slot passes, if the channel is still idle, the backoff count is decremented by one; if busy is detected during the countdown, the countdown is immediately frozen and the remaining count is preserved; the listening is maintained during the freezing; when the NAV = 0 and the CCA is idle and continuously remains for no less than the DIFS, the countdown is resumed, and the decrement continues from the preserved remaining count without re-extraction; the transmission right is obtained when the backoff count is decremented to zero, and the transmission stage is immediately entered; if the channel suddenly becomes busy at this time, it is considered as a conflict risk, and the gating is re-entered (only when the NAV timer = 0 and the CCA state = idle and continuously remains for no less than the DIFS);

[0122] The physical layer protocol data to be sent is sent immediately after obtaining the sending permission; the receiving end returns an acknowledgement frame after a short interframe interval when successfully receiving and completing the verification; the sending end judges as a failure event when the acknowledgement frame is not received within a preset acknowledgement timeout; specifically, the physical layer protocol data to be sent is immediately transmitted after obtaining the sending right; the sending start time and the frame length are recorded for subsequent statistics. Unicast: the receiving end returns an acknowledgement frame after a SIFS when successfully framing; the sending end listens to the acknowledgement within an ACK timeout timer. The acknowledgement is received to judge as success. Broadcast or control frame without acknowledgement: no acknowledgement is waited for; the sending is completed to judge as success, and the next backoff is directly entered; the acknowledgement frame is not received before the ACK timeout timer expires, the current sending is judged as a failure, and the retry processing is entered;

[0123] The extended interframe spacing silence is entered after the failure event occurs; the backoff flow process can be entered again after the extended interframe spacing timer expires and the channel satisfies the idle judgment condition; when the last failure is not obtaining the sending right and is not having sent unacknowledged, the remaining backoff count before freezing is used; when the last failure is having sent unacknowledged, the backoff count is extracted again and the competition window is updated according to the implementation strategy; the frame transmission is finally judged as a failure when the maximum retry number is reached; specifically, the EIFS timer silence is immediately entered when the failure is confirmed or obvious collision signs are detected; the backoff can be entered again only after the EIFS timer expires and the gating is satisfied. Two types of failure sources are distinguished and different processing is taken: Type A: not obtaining the sending right (interrupted in the countdown): not counted as "retry", not updating the CW, using the remaining backoff count before freezing, and continuing the countdown after the next gating is met. Type B: having sent unacknowledged: counted as "retry", increasing the competition window according to the implementation strategy (such as doubling expansion) but not exceeding the upper limit; resetting the backoff count and extracting again; the retry count is added by one, the maximum retry number is reached to judge as final failure and report. Once the acknowledgement frame is received and judged as success, the CW is reset to the minimum window, the retry count is cleared, and the next frame is re-entered from the gating.

[0124] The application is further configured to accumulate media access statistics in a sliding time window, including at least the number of attempts, the number of successes, the number of retransmissions, the number of acknowledgement timeout, the average backoff delay, the number of freezes and the length of extended interframe space silence; the media access statistics are time-stamped and configured with an identifier, and are aligned with the access layer statistics entries in the link quality indicator set. Specifically, the number of attempts is the number of actual initiations of transmission within the window (including the first transmission and retransmission), and broadcast is also counted as an attempt; the number of successes is the number of transmissions determined to be successful within the window; for unicast, "received acknowledgement frame" is used as the criterion, and for broadcast or control frames that do not require acknowledgement, "transmission completion" is used as the criterion; the number of retransmissions is the sum of the number of retransmissions triggered by the fact that the transmission has been sent but not confirmed (only for unicast); the number of acknowledgement timeouts is the number of events recorded within the window due to the fact that an acknowledgement frame has not been received within the acknowledgement timeout (only for unicast); the average backoff delay is the average of the cumulative length of the effective countdown from entering backoff to obtaining the right to transmit within the window; only the time when the countdown is idle and decreasing is counted, excluding the freeze period and the fixed interval between frames; the number of freezes is the cumulative number of freeze events within the window triggered by the fact that the channel has changed from idle to busy during the backoff countdown phase; the length of the extended interframe space silence is the total length of the extended interframe space silence state within the window (sum of all silence segments after failure).

[0125] The application is further configured to update the transmission quality parameters of the neighbor and the candidate path with the periodic routing advertisement packet, establish and maintain the routing table entry including the destination address, the next hop, the transmission quality value and the invalidation time, and execute the elimination on the entry which is not refreshed beyond the invalidation time; specifically, the data structure of the routing table entry and the timer setting include: the node identification and the interface information: the interface identification and the configuration identification are allocated for each wireless interface; the node identification is recorded; the neighbor table: the last received advertisement time, the local transmission quality observation of the link direction and the invalidation time are maintained according to the neighbor node; the routing candidate table: the multiple candidate next hops and the transmission quality values and the invalidation time thereof are saved according to the destination address as the key; the working routing table: the working next hop, the current transmission quality value and the minimum residence expiration time of each destination are recorded; the timer: the advertisement period timer, the neighbor and candidate aging timer, the working routing residence timer and the statistical window rolling timer. The routing advertisement packet and the processing include: the advertisement content: the source node, the advertisement serial number, the destination address entry and the transmission quality value, the path age or the recommended invalidation time and the advertisement generation time are included; the sending strategy: the advertisement is broadcasted according to the fixed period; the fast advertisement is triggered once after the working route switching; the receiving processing: according to each destination entry, the transmission quality value of the candidate path is formed for the destination in combination with the local link observation of "to the neighbor"; if the candidate does not exist, the candidate is added; if the candidate exists, the invalidation time is updated and refreshed; and the neighbor table entry is updated at the same time. The routing table maintenance and the timeout elimination include: the neighbor aging: the candidate with the next hop of the neighbor is deleted from the candidate table when the neighbor table entry is not refreshed beyond the invalidation time; the candidate aging: the candidate path is eliminated when the candidate path is not refreshed beyond the invalidation time; the working route checking: if the working next hop no longer exists in the candidate set or the candidate of the working next hop has expired, the working next hop is marked as "to be reevaluated" and enters the next hop reselection process.

[0126] The transmission quality values of the candidate next hops are compared on each destination address, and the working next hop is selected according to the maximum transmission quality criterion; specifically, all the candidates of the same destination are selected according to the principle of the maximum transmission quality value; if the new candidate is different from the current working next hop, the working next hop can be switched only when the minimum residence expiration time is expired and the transmission quality value of the new candidate exceeds the threshold advantage; otherwise, the status is maintained to avoid frequent jitter; the selected result is written into the working routing table, and the switching time and the switching reason are recorded;

[0127] The payload bit sequence is hop by hop forwarded according to the working next hop shown in the routing table; the table is repeatedly checked and forwarded in the relay node until the destination is reached; when the working next hop is missing or unreachable, path reevaluation is performed and forwarding is continued with the updated entry; specifically, after the upper layer delivers the payload bit sequence and the destination address, the working routing table is queried to obtain the working next hop; the link layer header and necessary control information are encapsulated according to the next hop; in the link layer sending, the access sending module is handed over to complete one-hop sending according to the idle judgment, backoff, sending and confirmation process; in the relay processing, the relay node repeatedly checks the table, encapsulates and sends after unpacking until the destination node; the broadcast service is directly sent out with the broadcast address, and no confirmation is maintained; when the maximum retry of the link layer is continuously reached and still not confirmed, or an explicit unreachable feedback is received, it is determined that the current hop fails and path reevaluation is entered; after the current hop fails, the candidate set for the destination is immediately reselected; if there is a suboptimal candidate and the gating condition is met (at least better than the current failed path), it is immediately switched and retransmitted; if there is no candidate available within a short time, the destination is marked as “temporarily unreachable”, the retry is delayed and new announcements are waited for; at the same time, the forwarding priority of the destination is reduced; the path failure count is counted, and when the threshold is exceeded, the preference for the next hop is temporarily reduced to suppress repeated selection of poor paths;

[0128] The end-to-end time stamp and hop-by-hop result of the sent and arrived packets are recorded, the end-to-end effective throughput, arrival success rate, delay and jitter, path switching frequency and approximate value of expected transmission number are calculated, and the statistical vector and time stamp are output within a sliding time window; specifically, the sending time stamp, byte number, selected working next hop and configuration identifier at path switching are recorded at the source node; the arrival confirmation and arrival time stamp are returned at the destination node or application layer (if there is no end-to-end confirmation, it can be replaced by the upper layer protocol or service receipt), including: end-to-end effective throughput: the payload byte amount of successfully arrived packets in the window divided by the window length; arrival success rate: the proportion of successfully arrived packets in the window; delay and jitter: the mean and fluctuation amplitude of the end-to-end delay of the packets in the window; path switching frequency: the number of working routing switching in the window; approximate value of expected transmission number: approximate statistics of “average sending attempt number for successful arrival” in the window; hop count record: the average hop count is recorded according to the hop count information carried by the last successful sample or announcement; the statistical vector and time stamp are output with the window start and end time as the boundary, and the interface identifier, destination classification and configuration identifier are carried;

[0129] The end-to-end statistics and media access statistics are aligned on the time window and bound to configuration identifiers, and then assembled together with the physical layer statistics into end-to-end and access layer entries of the link quality indicator set. Specifically, the end-to-end statistics and media access statistics within the same window are aligned according to "interface identifier, destination category, configuration identifier, and time window". If necessary, cross-window samples are merged or distributed, and the end-to-end statistics are written into the end-to-end layer entries of the link quality indicator set, and the media access statistics are written into the access layer entries. They are stored together with the physical layer statistics in the same set to form a hierarchical and traceable metric set.

[0130] Furthermore, such as Figures 4-7 As shown, since CSMA / CA adopts a contention-backoff channel selection method, in multi-hop application scenarios, the distance between Node 1 and Node 4 is considered to be far enough that the channel interference between them is negligible. Therefore, the channels of Node 1 and Node 4 can be reused without contention. Thus, in multi-hop application scenarios, after 4 hops, the channel is reused and the service rate no longer decreases. Theoretically, it can support the requirements of 255 nodes and 254 hops, and can meet the requirements of 64 network nodes and 63 hops.

[0131] The system supports point-to-point, point-to-multipoint, star, chain, and mesh networks to meet project requirements.

[0132] For example Figure 4 In the mesh network scenario shown, if node 2 needs to send a unicast service to node 5, it queries the routing table shown in the figure and finds that the route to the destination node 5 is directly through node 5. Therefore, it directly sends the unicast data to node 5.

[0133] If Node 2 has a broadcast service to send, it will send the broadcast data to the wireless channel. Since all other nodes are within one hop of Node 2, they can all receive the broadcast data sent by Node 2.

[0134] In addition, for the multi-hop network scenario shown in the figure below, if node 2 has a unicast service to send to node 5, then the routing table shown in the figure is queried and it is found that the route to the destination node 5 is node 4. That is, node 5 can be reached through node 4. Therefore, the unicast data is sent to the next hop node 4.

[0135] After receiving the unicast data, Node 4 finds that the destination node of the data is Node 5. It continues to query the routing table and finds that the route from this node to Node 5 is to reach Node 5 directly through Node 5. Therefore, it directly sends the unicast data to Node 5.

[0136] If node 2 has a broadcast service to send, it will send the broadcast data to the wireless channel. After receiving the broadcast data, nodes 3 and 4 will forward the broadcast data, and node 5 will then receive the forwarded broadcast data.

[0137] Embodiment two:

[0138] The embodiment of the MIMO-OFDM-based adaptive modulation anti-multipath UAV swarm communication system for realizing the above-mentioned MIMO-OFDM-based adaptive modulation anti-multipath UAV swarm communication method comprises:

[0139] The synchronization processing module receives the complex baseband sample sequence after the radio frequency down-conversion, denoted as a first data stream; performs coarse timing estimation on the first data stream to generate a coarse synchronization parameter set and a first compensated data stream; performs fine timing estimation on the first compensated data stream to generate a fine synchronization parameter set and a second compensated data stream;

[0140] The channel phase estimation module removes the cyclic prefix from the second compensated data stream symbol by symbol according to the fine synchronization parameter set and the system configuration set, and performs 64-point FFT to obtain a subcarrier frequency domain symbol matrix, denoted as a first frequency domain data; calculates a subcarrier-level initial channel estimation matrix and a noise power estimation according to the first frequency domain data; calculates a common phase error sequence and performs phase compensation based on the first frequency domain data and the noise power estimation to obtain a second frequency domain data;

[0141] The equalization decoding module performs frequency domain equalization and soft demapping on the second frequency domain data symbols with the subcarrier-level initial channel estimation matrix and the noise power estimation, and obtains a control parameter set through forward error correction decoding and CRC check; performs equalization decoding on the data symbols in the second frequency domain data according to the control parameter set to obtain a payload bit sequence, collects link quality parameters, and generates a link quality indication set;

[0142] The sender packaging module selects a target modulation and coding scheme, a target bandwidth, and a target spatial processing mode in the modulation and coding set according to the link quality indication set and the queue state, and forms a control triple; configures the sender mapping, power, and MIMO weight / codebook according to the control triple, and packages the payload bit sequence into a to-be-sent physical layer protocol data;

[0143] The access sending module sends the to-be-sent physical layer protocol data; returns an acknowledgement frame when successfully received, and retries according to the rules when failed; generates media access statistics;

[0144] The routing forwarding module selects a next hop according to a routing table and a transmission quality parameter at the ad hoc routing layer, and implements multi-hop forwarding on the payload bit sequence to form an end-to-end transmission; generates end-to-end statistics, which are fed back to the link quality indication set together with the media access statistics.

[0145] It should be noted that the MIMO-OFDM based adaptive modulation anti-multipath UAV swarm communication system provided by the above embodiment and the MIMO-OFDM based adaptive modulation anti-multipath UAV swarm communication method provided by the above embodiment belong to the same concept, wherein the specific manner in which each module and unit performs operations has been described in detail in the method embodiments, and will not be described here. The MIMO-OFDM based adaptive modulation anti-multipath UAV swarm communication system provided by the above embodiment can be applied in practice, and the above functions can be completed by different functional modules according to needs, that is, the internal structure of the system is divided into different functional modules to complete all or part of the functions described above, and this is not limited herein.

[0146] The above merely provides a specific implementation of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of changes or replacements within the technical scope disclosed by the present application, which should be encompassed within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. An adaptive modulation anti-multipath UAV swarm communication method based on MIMO-OFDM, characterized in that, include: The complex baseband sampling sequence after RF downconversion is received and denoted as the first data stream; Perform coarse timing estimation on the first data stream to generate a coarse synchronization parameter set and the first compensated data stream; Perform precise timing estimation on the first compensated data stream to generate a precise synchronization parameter set and a second compensated data stream; Based on the fine synchronization parameter set and the system configuration set, the cyclic prefix is ​​removed symbol by symbol from the second compensated data stream and a 64-point FFT is performed to obtain the subcarrier frequency domain symbol matrix, which is denoted as the first frequency domain data. The subcarrier-level initial channel estimation matrix and noise power estimation are calculated based on the first frequency domain data; the common phase error sequence is calculated based on the first frequency domain data and noise power estimation, and phase compensation is performed to obtain the second frequency domain data; Frequency domain equalization and soft demapping are performed on the second frequency domain data symbols using the subcarrier-level initial channel estimation matrix and noise power estimation. The control parameter set is obtained through forward error correction decoding and CRC check. Based on the control parameter set, equalization decoding is performed on the data symbols in the second frequency domain data to obtain the payload bit sequence. Link quality parameters are collected and a link quality indication set is generated. Based on the link quality indicator set and queue status, the target modulation and coding scheme, target bandwidth, and target spatial processing mode are selected from the modulation and coding set to form a control triplet; according to the control triplet, the transmitter mapping, power, and MIMO weights / codebook are configured, and the payload bit sequence is encapsulated into physical layer protocol data to be transmitted. Send the physical layer protocol data to be sent; return an acknowledgment frame upon successful reception, and retry according to the rules upon failure; Generate media access statistics; In the self-organizing network routing layer, the next hop is selected based on the routing table and transmission quality parameters, and the payload bit sequence is forwarded in multiple hops to form end-to-end transmission. Generate end-to-end statistics, which are then fed back into the link quality indicator set along with media access statistics.

2. The adaptive modulation anti-multipath UAV swarm communication method based on MIMO-OFDM according to claim 1, characterized in that, A sliding autocorrelation window with delays in multiples of 16 is set up. The delay correlation quantity and corresponding energy are calculated, and a detection metric is constructed, which is the ratio of the square of the delay correlation quantity amplitude to the square of the corresponding energy. Platform entry is determined based on a threshold and a minimum hold length, and the first timing position is determined within the entry interval based on the maximum correlation amplitude or energy criterion. The first frequency offset is obtained based on the ratio of the complex phase of the correlation quantity to the delay, and de-rotation compensation is performed on the first data stream. The output consists of a coarse synchronization parameter set composed of the first timing position and the first frequency offset, as well as the first compensated data stream. Using the local long preamble as a template, a sliding cross-correlation is performed on the first compensated data stream. The maximum correlation peak exceeding the threshold is searched in two adjacent intervals of length 64. The second timing position is determined by the position of the first peak. The second frequency offset is obtained based on the ratio of the phase difference between the two peak complex correlation values ​​to the length of 64 points. Rotation compensation is then performed on the first compensated data stream. The output is a fine synchronization parameter set composed of the second timing position and the second frequency offset, as well as the second compensated data stream.

3. The adaptive modulation anti-multipath UAV swarm communication method based on MIMO-OFDM according to claim 1, characterized in that, The system configuration set is loaded before the task starts and serves as a common prior for both the sending and receiving ends. It consists of a subset of restricted parameters, including: Physical layer fixed parameter subset: transmission system is MIMO-OFDM; fast Fourier transform number is 64; forward error correction coding type is convolutional code; maximum number of generator and receiver antennas does not exceed 4; peak-to-average power ratio suppression method is amplitude limiting; transmit power control and channel sounding functions are enabled; The physical layer switchable subsets include: subcarrier spacing values ​​of {78.125kHz, 156.25kHz, 312.5kHz}; cyclic prefix lengths of {16, 24, 32, 40, 48, 56, 63} and their pairs; modulation schemes of {BPSK, QPSK, 16QAM, 64QAM}; and physical channel bandwidths of {5MHz, 10MHz, 20MHz, 40MHz}. Spatial processing capability subset: The MIMO mode set is limited to spatial diversity, beamforming, and spatial multiplexing; Access layer timing prior subset: Media access control adopts CSMA / CA, and short frame intervals and time slot times are set.

4. The adaptive modulation anti-multipath UAV swarm communication method based on MIMO-OFDM according to claim 3, characterized in that, Using the second timing position in the fine synchronization parameter set as the symbol start reference, the cyclic prefix is ​​stripped symbol by symbol from the second compensated data stream according to the cyclic prefix length in the system configuration set, and the transformation is completed according to the 64-point fast Fourier transform size in the system configuration set to obtain the subcarrier frequency domain symbol matrix covering the preamble, signaling and data symbols, which is denoted as the first frequency domain data. At the same time, the symbol delimitation index is retained for subsequent processing. Based on the predefined training symbol positions and their transmission references in the system configuration set, linear least squares estimation is performed on the ratio of the received value to the reference value for each subcarrier, and time averaging is performed across multiple training symbols to obtain the subcarrier-level initial channel estimation matrix. Energy averaging is performed based on the set of empty subcarriers defined by the system configuration set, or energy estimation is performed based on the residual between the received value of the training symbol and the predicted value of the initial channel estimation. Either one can be selected or the values ​​can be combined to obtain the noise power estimate. Based on the pilot subcarrier positions and pilot symbols given in the system configuration set, and in conjunction with the initial channel estimation, a single common phase error is calculated for each orthogonal frequency division multiplexing symbol; when the pilot confidence is insufficient, inter-symbol smoothing or extrapolation is used to stabilize the common phase error. Phase compensation is performed on all subcarriers of each orthogonal frequency division multiplexing symbol by rotating them in the opposite direction of the common phase error. The compensated subcarrier frequency domain symbol matrix is ​​output and denoted as the second frequency domain data.

5. The adaptive modulation anti-multipath UAV swarm communication method based on MIMO-OFDM according to claim 1, characterized in that, At the control field symbol position, the subcarrier-level equalization weight is calculated according to the linear minimum mean square error criterion based on the subcarrier-level initial channel estimation matrix and noise power estimation to obtain the equalization symbol sequence of the control field; the equalization symbols are then softly demapped according to the bit log likelihood ratio based on the preset control field modulation and mapping rules, and deinterleaving and forward error correction decoding are completed. The control parameter set is obtained through cyclic redundancy check. The control parameter set includes at least the modulation order, coding rate, interleaving and mapping scheme, spatial stream number and stream mapping, scrambling code and frame structure information of the data field. At the symbol position in the data field, equalization is performed on the second frequency domain data subcarrier by subcarrier according to the spatial processing mode and modulation and coding scheme indicated by the control parameter set; when it is spatial diversity mode, diversity is performed to form a single-stream equivalent symbol; when it is spatial multiplexing mode, multi-stream detection weights are constructed according to the linear minimum mean square error criterion and stream-level symbol estimates are output; when it is beamforming mode, equalization is completed according to the single-stream equivalent channel. Based on the modulation and mapping rules determined by the control parameter set, the bit log likelihood ratio is generated for the equalized symbols, and deinterleaving and forward error correction decoding matched by code rate are performed. Then, descrambling and cyclic redundancy check are completed, and the payload bit sequence is output. In multi-stream scenarios, codeword assembly is completed according to the codeword and stream mapping relationship indicated by the control parameter set. During equalization, soft demapping and decoding, the physical layer parameters collected include: constellation error based on pilot or decision aids, mean or quantile statistics of bit log-likelihood ratio amplitude, block error rate or frame error rate, received signal strength or equivalent signal-to-noise ratio, single-hop effective throughput measured by window, and noise power and common phase drift amplitude. Physical layer parameter statistics are aggregated by field and subcarrier into a set of link quality indicators.

6. The adaptive modulation anti-multipath UAV swarm communication method based on MIMO-OFDM according to claim 1, characterized in that, Under the joint constraints of the link quality indicator set and the queue status, the target modulation and coding scheme, target bandwidth, and target spatial processing mode are determined in the modulation and coding set, the bandwidth set, and the spatial processing mode set, respectively. The target modulation and coding scheme is based on the block error rate target before decoding and the equivalent signal-to-noise ratio threshold, and hysteresis and minimum dwell period are set. The target bandwidth is based on the busy ratio, retransmission rate, or queue backlog threshold and is synchronized with the target modulation and coding scheme. The target spatial processing mode is based on the channel rank and correlation metric and is constrained by noise power estimation and block error rate. Based on the target modulation and coding scheme, set the modulation order, coding rate, interleaving and bit mapping rules; based on the target spatial processing mode, determine the number of spatial streams, codeword-to-stream mapping and combination or multiplexing decoupling strategies; based on the target bandwidth, configure subcarrier occupancy and pilot patterns. Within the spectral mask and maximum power generation limit, the transmit power plan is determined based on the error constraints of the target receiver signal-to-noise ratio and the link quality indicator set, and peak-to-average power ratio suppression is performed on the symbol sequence to be transmitted to reduce nonlinear distortion; When the target spatial processing mode is spatial diversity, load the diversity merging or space-time coding weights; when it is spatial multiplexing, select or update the precoding matrix according to the subcarrier-level initial channel estimation matrix and allocate the power of each stream; when it is beamforming, select or calculate the beam weight vector from the codebook and generate the corresponding index. The payload bit sequence is forward error correction encoded according to the target modulation and coding scheme and cyclic redundancy check is added. Interleaving and bit-to-constellation mapping are performed. Pilots and guard subcarriers are inserted according to the target bandwidth and target space processing mode and the flow-by-flow symbol assembly is completed. Assemble the preamble and training fields, signaling or control fields, and data fields to obtain the physical layer protocol data to be transmitted. The signaling or control fields explicitly carry the target modulation and coding scheme, target bandwidth and target spatial processing mode, codebook or precoding index, frame length, and configuration identifier.

7. The adaptive modulation anti-multipath UAV swarm communication method based on MIMO-OFDM according to claim 1, characterized in that, The system employs a combination of physical carrier sensing and virtual carrier sensing to determine idle status. Virtual carrier sensing updates the network allocation vector timer based on the frame duration field. The system only enters the backoff process when the network allocation vector is zero and physical carrier sensing determines that the system is idle and maintains an idle status for at least the DCF frame interval. Randomly select the backoff count as the initial countdown value within the current contention window; decrement the countdown in each time slot; freeze the countdown and retain the remaining count when the channel is busy; resume the countdown when the channel is idle again and the countdown is maintained for at least the DCF frame interval; obtain transmission permission when the backoff count decreases to zero. Upon obtaining permission to send, the physical layer protocol data to be sent is sent immediately; upon successful reception and completion of verification, the receiving end returns an acknowledgment frame after a short frame interval; if the sending end does not receive an acknowledgment frame within a preset acknowledgment timeout, it is considered a failure event. After a failure event occurs, the system enters a period of extended frame interval silence. The backoff process can only be re-entered after the extended frame interval timer expires and the channel meets the idle determination condition. If the previous failure was due to not obtaining the right to send and not having sent but not acknowledged, the remaining backoff count before freezing is used. If the previous failure was due to having sent but not acknowledged, the backoff count is recalculated and the contention window is updated according to the implementation strategy. When the maximum number of retries is reached, the transmission of the frame is determined to have ultimately failed.

8. The adaptive modulation anti-multipath UAV swarm communication method based on MIMO-OFDM according to claim 7, characterized in that, The media access statistics are accumulated within the sliding time window, including at least the number of attempts, the number of successful attempts, the retransmission count, the confirmation timeout count, the average backoff delay, the number of freezes, and the extended frame interval silence duration. Media access statistics are timestamped and have configuration identifiers, and are aligned with access stratum statistics entries in the link quality indicator set.

9. The adaptive modulation anti-multipath UAV swarm communication method based on MIMO-OFDM according to claim 1, characterized in that, The system updates the transmission quality parameters of neighbors and candidate paths with periodic route advertisement messages, establishes and maintains routing table entries, including destination address, next hop, transmission quality value and expiration time, and evicts entries that have not been updated after the expiration time. At each destination address, the transmission quality values ​​of the candidate next hops are compared, and the working next hop is selected based on the maximum transmission quality criterion. The payload bit sequence is forwarded hop-by-hop according to the working next hop shown in the routing table; Repeatedly look up and forward at relay nodes until the destination is reached; when the next hop is missing or unreachable, perform path re-evaluation and continue forwarding with the updated entry. Record end-to-end timestamps and hop-by-hop results for sent and arrived packets, calculate end-to-end effective throughput, arrival power, delay and jitter, path switching frequency and approximate values ​​of expected transmission count, and output statistical vectors and timestamps within a sliding time window. Align end-to-end statistics and media access statistics on a time window and bind configuration identifiers, then assemble them together with physical layer statistics into end-to-end and access layer entries for the link quality indicator set.

10. A MIMO-OFDM-based adaptive modulation anti-multipath UAV swarm communication system, used to implement the MIMO-OFDM-based adaptive modulation anti-multipath UAV swarm communication method according to any one of claims 1-9, characterized in that, include: Synchronization processing module: Receives the complex baseband sampling sequence after RF downconversion, denoted as the first data stream; Perform coarse timing estimation on the first data stream to generate a coarse synchronization parameter set and the first compensated data stream; Perform precise timing estimation on the first compensated data stream to generate a precise synchronization parameter set and a second compensated data stream; Channel phase estimation module: Based on the fine synchronization parameter set and system configuration set, remove the cyclic prefix symbol by symbol from the second compensated data stream and perform a 64-point FFT to obtain the subcarrier frequency domain symbol matrix, which is denoted as the first frequency domain data; The subcarrier-level initial channel estimation matrix and noise power estimation are calculated based on the first frequency domain data; the common phase error sequence is calculated based on the first frequency domain data and noise power estimation, and phase compensation is performed to obtain the second frequency domain data; The equalization decoding module performs frequency domain equalization and soft demapping on the second frequency domain data symbols using the subcarrier-level initial channel estimation matrix and noise power estimation. After forward error correction decoding and CRC check, the control parameter set is obtained. Based on the control parameter set, equalization decoding is performed on the data symbols in the second frequency domain data to obtain the payload bit sequence. Link quality parameters are collected, and a link quality indication set is generated. Transmitter encapsulation module: Based on the link quality indicator set and queue status, selects the target modulation and coding scheme, target bandwidth, and target spatial processing mode within the modulation and coding set to form a control triplet; configures the transmitter mapping, power, and MIMO weights / codebook according to the control triplet, and encapsulates the payload bit sequence into physical layer protocol data to be transmitted; Access transmission module: transmits the physical layer protocol data to be sent; returns an acknowledgment frame upon successful reception, and retryes according to the rules upon failure; Generate media access statistics; Routing and forwarding module: In the self-organizing network routing layer, the next hop is selected based on the routing table and transmission quality parameters, and the payload bit sequence is forwarded in multiple hops to form end-to-end transmission; Generate end-to-end statistics, which are then fed back into the link quality indicator set along with media access statistics.

Images