Distributed precoding system and method of asynchronous cellular-free wireless network

Abstract

The invention discloses a distributed precoding system and method for an asynchronous cellular-free wireless network, and the system comprises radio frequency access points (AP) of all frequency bands, distributed processing units, and needed wired connections and switches, and the APs are distributed in a wide area. The AP mainly completes wireless radio frequency signal sending and receiving of each frequency band, digital-to-analog and analog-to-digital conversion and waveform signal processing of a physical channel. The distributed processing unit mainly completes channel parameter estimation, downlink precoding calculation and precoding. According to the invention, the problem of existing non-cellular large-scale MIMO asynchronous interference can be solved in a distributed precoding mode.

Description

Technical Field

[0001] The present invention relates to the field of wireless communication technology, and in particular to a distributed precoding system and method for an asynchronous non-cellular wireless network. Background Art

[0002] Cellular networks include macro cells with wide-area coverage and micro cells and pico cells with local coverage. Dense network layout can significantly improve network throughput, but the edge user problem has never been effectively solved. Early distributed antenna systems can and cannot solve this problem well. Cell-free networks, as a distributed implementation version of massive multiple-input multiple-output (MIMO) antennas, make it possible to solve the interference problem of cell edge users. In cell-free wireless access networks, distributed signal processing can realize cell-free massive MIMO cooperative transmission.

[0003] However, in terms of implementation, first of all, there is a general trade-off between the distribution capability and spectrum efficiency of non-cellular networks: fully centralized processing can effectively suppress interference and noise to achieve a throughput that is difficult to achieve in traditional cellular networks, but it places extremely high computing power requirements on the central processing unit and requires a large amount of backhaul cost in each time-frequency resource block; distributed downlink processing can disperse the calculation to distributed baseband units or even wireless radio frequency units, greatly reducing the cost of backhaul, but the current distributed downlink processing performance is difficult to compare with the fully centralized method. Another important problem to be solved by the present invention is asynchronous transmission. Current research on non-cellular networks often assumes that the system is perfectly synchronized, but in fact, in a wide-coverage non-cellular network, the probability of asynchronous conditions is quite high. The large propagation distance difference causes the cyclic prefix to be unable to cover the seriously misaligned signal, causing asynchronous interference. Summary of the invention

[0004] Purpose of the invention: The present invention provides a distributed precoding system and method for an asynchronous non-cellular wireless network, which solves the existing non-cellular large-scale MIMO asynchronous interference problem in a distributed precoding manner.

[0005] Technical solution: The distributed precoding system of an asynchronous non-cellular wireless network described in the present invention includes: radio frequency access points AP of each frequency band, distributed processing units and required wired connections and switches, and the APs are scattered in a wide area;

[0006] The AP mainly completes the transmission and reception of wireless RF signals in each frequency band, digital-to-analog and analog-to-digital conversion, and waveform signal processing of physical channels;

[0007] The distributed processing unit mainly completes channel parameter estimation, downlink precoding calculation and precoding.

[0008] Furthermore, the AP reports all real-time information to the corresponding distributed processing unit; the AP needs to obtain the delay from the served user to the AP in advance, and send the delay information to the corresponding distributed processing unit; the delay from the user to the AP is calculated by the distance from the user to the AP, and the distance is obtained based on the user location information.

[0009] Furthermore, the distributed processing units are connected by wires, and the distributed processing units exchange instantaneous or statistical information through direct connections or switches.

[0010] Furthermore, the information sent by the AP to the distributed processing unit includes the received user uplink sounding reference signal. The distributed processing unit aggregates the uplink sounding reference signals of the served users to perform channel estimation and calculates the downlink precoding using the reciprocity of the uplink and downlink channels.

[0011] Furthermore, the distributed processing unit summarizes the delays between all the managed APs and each user, finds the nearest AP for each user, and calculates the delay difference between other APs to the user and the nearest AP, and calculates the asynchronous transmission parameters based on the channel information and the delay difference.

[0012] Furthermore, in the channel estimation stage, the system uses a long cyclic prefix to offset the asynchronous effect and estimates a channel without asynchronous influence. In the downlink transmission stage, the system uses a short cyclic prefix to improve the spectral efficiency. At this time, asynchronous precoding is required to suppress the impact of the generated asynchronous interference.

[0013] Accordingly, a distributed precoding method for an asynchronous non-cellular wireless network includes the following steps:

[0014] Step 1: AP first measures the distance of the users it serves, and each AP obtains the local air interface propagation time set Secondly, the AP receives the uplink pilot signal sent by the user and transmits it back to the corresponding distributed processing unit together with the air interface propagation time set;

[0015] Step 2: The distributed processing unit summarizes the uplink detection pilot signal transmitted by the AP and performs channel estimation, summarizes the air interface propagation time set transmitted by the AP, and calculates the downlink asynchronous transmission parameters in combination with the estimated channel calculation;

[0016] Step 3: The distributed processing unit performs local precoding calculation based on the estimated channel and the calculated asynchronous transmission parameter function. Then, the distributed processing unit calculates the equivalent channel and asynchronous parameter information after being multiplied by the precoding product, and shares this information with other distributed units. Finally, the distributed processing unit optimizes the local precoding based on the information shared by other distributed processing units.

[0017] Furthermore, in step 1, the AP first performs ranging for the users it serves, and each AP obtains the local air interface propagation time set Secondly, the AP receives the uplink pilot signal sent by the user, and transmits the air interface propagation time set back to the corresponding distributed processing unit, which specifically includes the following steps:

[0018] Step 101: The AP measures the user distance using the flight time method. For example, the user initiates a measurement signal. After receiving the distance measurement signal, the AP obtains the air interface propagation time based on the sending start time recorded in the signal and the time when the AP receives the signal, and then deducts the hardware processing time.

[0019] Step 102: In the uplink channel estimation phase, the system uses a long cyclic prefix or long zero padding to avoid asynchronous interference of the uplink pilot signal. The user sends an uplink pilot signal to the AP at the same time according to the allocated pilot. Due to the existence of the long cyclic prefix, the pilot signal received by the AP has no asynchronous interference.

[0020] Step S103: In the uplink channel estimation phase, the AP receives the uplink detection pilot signals from the users and directly transmits them back to the distributed processing unit without processing.

[0021] Further, in step 2, the distributed processing unit summarizes the uplink detection pilot signal returned by the AP and performs channel estimation, and also summarizes the air interface propagation time set returned by the AP, and calculates the downlink asynchronous transmission parameter in combination with the estimated channel, specifically including the following steps:

[0022] Step 201: The distributed processing unit aggregates the uplink detection pilot signals received by the managed APs and performs channel estimation. According to the uplink and downlink reciprocity, the downlink precoding is calculated. Assume that the distributed processing unit estimates a perfect channel, that is, the distributed processing unit has the information set If the mth AP is connected to the distributed processing unit;

[0023] Step 202: The distributed processing unit summarizes the distances between all the managed APs and each user, finds the distance to the nearest AP for each user, and calculates the difference between the air interface propagation time from other APs to the user and the air interface propagation time of the nearest AP.

[0024]

[0025] According to the channel information and the air interface propagation time difference, the asynchronous transmission parameters are calculated, including the coefficient a of the inter-carrier interference (ICI) n,n′,k′ and the coefficient b of Inter-Symbol-Interference (ISI) n,n′,k′The meanings of these two symbols are ICI / ISI caused by the nth subcarrier to the kth user on the n'th subcarrier; and the equivalent channel for transmitting the desired signal for the k'th user on the n'th subcarrier is c n′,k′ To express.

[0026] Further, in step 3, the distributed processing unit performs local precoding calculation according to the estimated channel and the calculated asynchronous transmission parameter function, and then the distributed processing unit calculates the equivalent channel and asynchronous parameter information after being multiplied by the precoding product, and shares this information with other distributed units. Finally, the distributed processing unit optimizes the local precoding according to the information shared by other distributed processing units, specifically including the following steps:

[0027] Step 301: In the downlink transmission phase, the system uses a short cyclic prefix. At this time, there is asynchronous interference. The local precoding in step 3 should take this asynchronous interference into consideration for precoding design. The problem established is as follows:

[0028]

[0029] By optimizing the above problems, the multi-user interference and asynchronous interference of the local distributed processing unit are suppressed, but the multi-user interference and asynchronous interference of the subnets managed by the distributed processing units still exist;

[0030] Step 302: The distributed processing unit pre-equalizes the local channel and asynchronous interference parameters using the local precoding obtained in step 301. The dimensions of the equivalent channel and the equivalent asynchronous parameters after pre-equalization are much smaller than the original channel and the asynchronous parameters, so that the amount of information to be shared is greatly reduced. The distributed processing unit can adjust its sharing strategy in real time according to the capacity of the fronthaul link, that is, share its information in real time or statistically.

[0031] Step S303: The distributed processing unit optimizes local precoding according to the information shared by other distributed processing units, summarizes local and received channel and asynchronous parameters, and solves a K×K matrix to optimize local precoding. The optimization matrix balances multi-user interference and asynchronous interference between distributed processing units.

[0032] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages: the asynchronous team precoding proposed in the present invention has a mathematical analytical expression and is supported by team theory as the underlying principle, without any iterative calculation, and because of the adoption of the team decision theory framework, asynchronous precoding can be performed in a distributed manner; in addition, the present invention can also control the fronthaul capacity to achieve the optimal compromise between performance and cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1It is a schematic diagram of the system structure of the present invention.

[0034] Figure 2 It is a schematic diagram of the method flow of the present invention.

[0035] Figure 3 This is a Monte Carlo simulation result diagram of the asynchronous distributed precoding of the present invention, which compares the cumulative distribution of user rates of minimum mean square error (MMSE) precoding without and with asynchronous influence, and the maximum signal-to-leakage-and-noise (SLNR) with asynchronous influence taken into account.

[0036] Figure 4 This is a Monte Carlo simulation result diagram of the asynchronous distributed precoding of the present invention, which compares the cumulative distribution of user rates of the asynchronous distributed precoding proposed by the present invention under different fronthaul capacities. DETAILED DESCRIPTION

[0037] A distributed precoding implementation method for an asynchronous non-cellular wireless network is provided to solve the asynchronous transmission problem in the non-cellular network, and the method can be implemented in a distributed manner to meet the requirements of distributed deployment of the non-cellular network.

[0038] For ease of description, M represents the number of APs, K represents the number of users, N represents the number of subcarriers, T represents the duration of OFDM symbols, and the length of the cyclic prefix is ​​T. CP The total duration of the OFDM symbol and the cyclic prefix is ​​represented by T s To express, the maximum length of the channel impulse response is T D In addition, T m,k represents the air interface propagation time from the mth AP to the kth user; Δ m,k This means that the kth user has an extra air interface propagation time from the mth AP compared to the k'th AP that is closest to it, that is, Δ m,k =T m,k -T m,k′ At the same time, the end time of the ath OFDM symbol sent by the mth AP and received by the kth user is recorded as Assuming that each AP has a single antenna, the channel between the mth mAP and the kth user is

[0039]

[0040] Among them, β m,k is large-scale fading, H m,n,k This is a small-scale fading, but the present invention is not limited to a single-antenna AP.

[0041] The system implementation of the present invention can be divided into three parts:

[0042] 100 is AP measurement and backhaul;

[0043] 200 is a distributed processing unit for parameter calculation;

[0044] 300 is a distributed processing unit that performs asynchronous distributed precoding locally and collaboratively.

[0045] In this embodiment, it is divided into three steps:

[0046] Step S1: AP first performs ranging (or propagation delay measurement) on the users it serves, and each AP obtains the local air interface propagation time set Secondly, the AP receives the uplink pilot signal sent by the user and transmits it back to the corresponding distributed processing unit together with the air interface propagation time set.

[0047] The specific steps include:

[0048] Step S101, optionally, the AP can use the flight time method to measure the user distance, such as when the user initiates a measurement signal, and after the AP receives the distance measurement signal, it can obtain the air interface propagation time based on the sending start time recorded in the signal, combined with the time when the AP receives the signal, and then deducting the hardware processing time. However, the present invention is not limited to any distance measurement method.

[0049] Step S102: In the uplink channel estimation phase, the system uses a long cyclic prefix or long zero padding to avoid asynchronous interference of the uplink pilot signal. The user sends an uplink pilot signal to the AP at the same time according to the allocated pilot. Due to the existence of the long cyclic prefix, the pilot signal received by the AP has no asynchronous interference.

[0050] Step S103: In the uplink channel estimation phase, the AP receives the uplink detection pilot signals from the users and directly transmits them back to the distributed processing unit without processing.

[0051] Step S2: The distributed processing unit aggregates the uplink detection pilot signal returned by the AP and performs channel estimation, aggregates the air interface propagation time set returned by the AP, and calculates the downlink asynchronous transmission parameters in combination with the estimated channel.

[0052] The specific steps include:

[0053] Step S201: The distributed processing unit aggregates the uplink detection pilot signals received by the managed APs and performs channel estimation, and calculates the downlink precoding according to the uplink and downlink reciprocity. Channel estimation methods include but are not limited to least squares estimation (LS), minimum mean square error estimation (MMSE), and other linear and nonlinear existing multi-antenna channel estimation methods. In this embodiment, in order to simplify the representation, it can be assumed that the distributed processing unit estimates a perfect channel, that is, the distributed processing unit has an information set If the mth AP is connected to the distributed processing unit.

[0054] Step S202: The distributed processing unit summarizes the distances between all managed APs and each user, finds the distance to the nearest AP for each user, and calculates the difference between the air interface propagation time from other APs to the user and the air interface propagation time of the nearest AP.

[0055]

[0056] According to the channel information and the air interface propagation time difference, the asynchronous transmission parameters are calculated, including the coefficient a of the inter-carrier interference (ICI) n,n′,k′ and the coefficient b of Inter-Symbol-Interference (ISI) n,n′,k′ The meanings of these two symbols are the ICI / ISI caused by the nth subcarrier to the kth user on the n'th subcarrier. The equivalent channel for transmitting the desired signal for the k'th user on the n'th subcarrier is represented by c n′,k′ To express.

[0057] Step S3, the distributed processing unit performs local precoding calculation according to the estimated channel and the calculated asynchronous transmission parameter function, and then calculates the equivalent channel, asynchronous parameters and other information after the precoding product is multiplied, and shares this information with other distributed units. Finally, the distributed processing unit optimizes the local precoding according to the information shared by other distributed processing units.

[0058] The specific steps include:

[0059] Step S301: In the downlink transmission phase, the system uses a short cyclic prefix. At this time, there is asynchronous interference. The local precoding in step S3 should take this asynchronous interference into consideration for precoding design. The established problem is as follows:

[0060]

[0061] By optimizing the above problems, the multi-user interference and asynchronous interference of the local distributed processing unit are suppressed, but the multi-user interference and asynchronous interference of the subnets managed by the distributed processing units still exist. The local precoding methods include but are not limited to maximum ratio transmission, regularized zero forcing, maximum signal to interference and noise ratio, maximum signal to leakage and noise ratio and other methods.

[0062] Step S302: The distributed processing unit pre-equalizes the local channel and asynchronous interference parameters using the local precoding obtained in step S301, wherein the dimensions of the equivalent channel and the equivalent asynchronous parameters after pre-equalization are much smaller than the original channel and the asynchronous parameters, so that the amount of information shared is greatly reduced, and the distributed processing unit can adjust its sharing strategy in real time according to the capacity of the fronthaul link, that is, share its own information in real time or statistically.

[0063] Step S303, the distributed processing unit optimizes the local precoding according to the information shared by other distributed processing units. Typically, the local and received channel and asynchronous parameters are summarized to solve a K×K matrix to optimize the local precoding (which can be proved by team decision theory). The optimization matrix balances the multi-user interference and asynchronous interference between the distributed processing units.

[0064] Through the above three major steps, the distributed precoding of asynchronous non-cellular wireless network is completed, which is very close to the effect when there is no asynchronous interference, and the precoding is realized in a distributed way. The effect example can be seen Figure 3 and Figure 4 . Figure 3 This shows the necessity of asynchronous precoding, compared to the situation without asynchrony ( Figure 3 "SynchronousMMSE" in the MMSE), the performance is greatly degraded when there is asynchronous interference ( Figure 3 middle" MMSE”), but considers the case where asynchronous SLNR precoding of asynchronous interference terms can bring the system performance almost back to synchronization. Figure 4 The performance of the method of the present invention under different fronthaul capacity (i.e., the instantaneous amount of information that can be exchanged by distributed devices) configurations is demonstrated, and the performance of asynchronous distributed precoding with the ratio of the exchanged instantaneous information to the total exchanged information ranging from 0 to 1 is compared.

Claims

1. A distributed precoding system for an asynchronous non-cellular wireless network, characterized in that: include: RF access points (APs) for each frequency band, distributed processing units, and required wired connections and switches, with APs spread over a wide area; The AP mainly completes the transmission and reception of wireless RF signals in each frequency band, digital-to-analog and analog-to-digital conversion, and waveform signal processing of physical channels; The distributed processing unit mainly completes channel parameter estimation, downlink precoding calculation and precoding.

2. The distributed precoding system for asynchronous non-cellular wireless network as claimed in claim 1, characterized in that: The AP reports all real-time information to the corresponding distributed processing unit. The AP needs to obtain the delay from the user it serves to the AP in advance and send the delay information to the corresponding distributed processing unit. The delay from the user to the AP is calculated by the distance from the user to the AP, and the distance is obtained according to the user location information.

3. The distributed precoding system for asynchronous non-cellular wireless network as claimed in claim 1, characterized in that: The distributed processing units are connected by wires, and the distributed processing units exchange instantaneous or statistical information through direct connections or switches.

4. The distributed precoding system for asynchronous non-cellular wireless network as claimed in claim 1, characterized in that: The information sent by the AP to the distributed processing unit includes the received user uplink sounding reference signal. The distributed processing unit aggregates the uplink sounding reference signals of the served users to perform channel estimation and calculate the downlink precoding using the reciprocity of the uplink and downlink channels.

5. The distributed precoding system for asynchronous non-cellular wireless network as claimed in claim 1, characterized in that: The distributed processing unit summarizes the delays between all the managed APs and each user, finds the nearest AP for each user, and calculates the delay difference between other APs to the user and the nearest AP, and calculates the asynchronous transmission parameters based on the channel information and the delay difference.

6. The distributed precoding system of the asynchronous non-cellular wireless network as claimed in claim 1, characterized in that: In the channel estimation stage, the system uses a long cyclic prefix to offset the asynchronous effect and estimates a channel without asynchronous influence. In the downlink transmission stage, the system uses a short cyclic prefix to improve the spectrum efficiency. At this time, asynchronous precoding is required to suppress the impact of the generated asynchronous interference.

7. A method for a distributed precoding system applied to an asynchronous non-cellular wireless network as claimed in claim 1, characterized in that: The steps include: Step 1: AP first measures the distance of the users it serves, and each AP obtains the local air interface propagation time set Secondly, the AP receives the uplink pilot signal sent by the user and transmits it back to the corresponding distributed processing unit together with the air interface propagation time set; Step 2: The distributed processing unit summarizes the uplink detection pilot signal transmitted by the AP and performs channel estimation, summarizes the air interface propagation time set transmitted by the AP, and calculates the downlink asynchronous transmission parameters in combination with the estimated channel calculation; Step 3: The distributed processing unit performs local precoding calculation based on the estimated channel and the calculated asynchronous transmission parameter function. Then, the distributed processing unit calculates the equivalent channel and asynchronous parameter information after being multiplied by the precoding product, and shares this information with other distributed units. Finally, the distributed processing unit optimizes the local precoding based on the information shared by other distributed processing units.

8. The distributed precoding method of the asynchronous non-cellular wireless network as claimed in claim 7, characterized in that: In step 1, the AP first measures the distance of the served users, and each AP obtains the local air interface propagation time set Secondly, the AP receives the uplink pilot signal sent by the user, and transmits the air interface propagation time set back to the corresponding distributed processing unit, which specifically includes the following steps: Step 101: The AP measures the user distance using the flight time method. For example, the user initiates a measurement signal. After receiving the distance measurement signal, the AP obtains the air interface propagation time based on the sending start time recorded in the signal and the time when the AP receives the signal, and then deducts the hardware processing time. Step 102: In the uplink channel estimation phase, the system uses a long cyclic prefix or long zero padding to avoid asynchronous interference of the uplink pilot signal. The user sends an uplink pilot signal to the AP at the same time according to the allocated pilot. Due to the existence of the long cyclic prefix, the pilot signal received by the AP has no asynchronous interference. Step S103: In the uplink channel estimation phase, the AP receives the uplink detection pilot signals from the users and directly transmits them back to the distributed processing unit without processing.

9. The distributed precoding method of an asynchronous non-cellular wireless network as claimed in claim 7, characterized in that: In step 2, the distributed processing unit summarizes the uplink detection pilot signal returned by the AP and performs channel estimation, summarizes the air interface propagation time set returned by the AP, and calculates the downlink asynchronous transmission parameter in combination with the estimated channel. Specifically, the steps include: Step 201: The distributed processing unit aggregates the uplink detection pilot signals received by the managed APs and performs channel estimation. According to the uplink and downlink reciprocity, the downlink precoding is calculated. Assume that the distributed processing unit estimates a perfect channel, that is, the distributed processing unit has the information set If the mth AP is connected to the distributed processing unit; Step 202: The distributed processing unit summarizes the distances between all the managed APs and each user, finds the distance to the nearest AP for each user, and calculates the difference between the air interface propagation time from other APs to the user and the air interface propagation time of the nearest AP. According to the channel information and the air interface propagation time difference, the asynchronous transmission parameters are calculated, including the coefficient a of the inter-carrier interference (ICI) n,n′,k′ and the coefficient b of Inter-Symbol-Interference (ISI) n,n′,k′ The meanings of these two symbols are ICI / ISI caused by the nth subcarrier to the kth user on the n'th subcarrier; and the equivalent channel for transmitting the desired signal for the k'th user on the n'th subcarrier is c n′,k′ To express.

10. The distributed precoding method of asynchronous non-cellular wireless network according to claim 7, characterized in that: In step 3, the distributed processing unit performs local precoding calculation according to the estimated channel and the calculated asynchronous transmission parameter function, and then the distributed processing unit calculates the equivalent channel and asynchronous parameter information after the precoding product is multiplied, and shares this information with other distributed units. Finally, the distributed processing unit optimizes the local precoding according to the information shared by other distributed processing units, which specifically includes the following steps: Step 301: In the downlink transmission phase, the system uses a short cyclic prefix. At this time, there is asynchronous interference. The local precoding in step 3 should take this asynchronous interference into consideration for precoding design. The problem established is as follows: By optimizing the above problems, the multi-user interference and asynchronous interference of the local distributed processing unit are suppressed, but the multi-user interference and asynchronous interference of the subnets managed by the distributed processing units still exist; Step 302: The distributed processing unit pre-equalizes the local channel and asynchronous interference parameters using the local precoding obtained in step 301. The dimensions of the equivalent channel and the equivalent asynchronous parameters after pre-equalization are much smaller than the original channel and the asynchronous parameters, so that the amount of information shared is greatly reduced. The distributed processing unit adjusts its sharing strategy in real time according to the capacity of the fronthaul link, that is, shares its information in real time or statistically. Step S303: The distributed processing unit optimizes local precoding according to the information shared by other distributed processing units, summarizes local and received channel and asynchronous parameters, and solves a K×K matrix to optimize local precoding. The optimization matrix balances multi-user interference and asynchronous interference between distributed processing units.

Images