WO2015154396A1 - Distributed system and closed-loop phase synchronization method therefor based on directional negative feedback - Google Patents

Abstract

The present application is applicable in the technical field of radio communications. Provided are a distributed system and a closed-loop phase synchronization method therefor based on directional negative feedback. With the present application, a large step-size disturbance is employed in several initial time slots when a received signal strength is less than a next-stage threshold, thus allowing rapid growth of the received signal strength of a receiver; a small step-size disturbance is employed in later-stage time slots when the received signal strength is less than a next-stage threshold, thus allowing for acquisition of the stability of the prior art, and iteratively allowing the signal strength of the receiver to converge to an ideal situation; also, when making adjustments, the direction of a disturbance step size of a transmitter is determined first, when the direction is determined, an adjustment direction for each transmitter is shown to be correct, only the magnitude of the disturbance step size is adjusted in subsequent time slots, thus allowing further accelerated convergence of subsequent received signal strengths.

Description

Distributed system and its closed-loop phase synchronization method based on directed negative feedback Technical field

The invention belongs to the field of wireless communication technologies, and in particular relates to a distributed system and a closed-loop phase synchronization method based on directed negative feedback.

Background technique

The single-bit feedback algorithm is a common technique for achieving transmitter phase synchronization at the receiving end. The latest result of the algorithm is a hybrid single-bit feedback algorithm based on distributed beamforming technology, which uses the -1 bit information fed back from the receiver to perform a step disturbance δ _i for each transmitter at the transmitting end ( n) phase compensation feedback receiver and a continuous number C _N "-1" when the bit information for a scale factor ε _i (n) attenuated δ _i (n), phase compensation can be timely adjustment of the transmitter. The advantage of this technique is that the received signal strength can be rapidly increased by phase compensation, and the phase attenuation can cause the received signal strength to converge in a certain time slot.

However, according to the research, the transmitter has different convergence speeds under different step sizes, and the convergence gain of the algorithm will be different under different feedback adjustment mechanisms. Therefore, the convergence speed and convergence ability of the existing algorithms have room for improvement. .

Summary of the invention

The first technical problem to be solved by the present invention is to provide a closed-loop phase synchronization method based on directional negative feedback for a distributed system, which aims to accelerate the convergence speed and convergence capability of the received signal strength of the receiver.

The present invention is implemented in a distributed system based on a closed-loop phase synchronization method with directed negative feedback, comprising the following steps performed in each time slot:

Step A: The receiver determines whether the strength of the received signal in the current time slot is greater than or equal to the threshold of the next stage, and feeds back corresponding single-bit information to the transmitter according to the determination result;

Step B: After the transmitter receives the single-bit information, it performs parsing. If the optimal direction of the disturbance step is not determined, and the analysis result is that the strength of the received signal of the receiver is greater than or equal to the threshold of the next stage, step C is performed, otherwise execution is performed. Step D; if the optimal direction of the disturbance step is determined, and the analysis result is that the strength of the received signal of the receiver is greater than or equal to the threshold of the next stage, step C is performed, otherwise step E is performed; the optimal direction of the disturbance step is When the received signal strength is less than the next-stage threshold, when the first received signal strength is greater than or equal to the next-stage threshold, the disturbance step direction of the previous time slot is the optimal direction of the disturbance step of all subsequent time slots;

Step C, the best phase of the transmitter in the next time slot still takes its best phase in the current time slot, and does not generate a disturbance step in the next time slot, and directly adds the initial phase to the best phase of the next time slot in the next time slot. Transmitting a signal as a phase of transmission, and adding one to the number of stages; the initial phase is obtained by estimating, based on a signal received by the transmitter, the transmitter in the first time slot;

In step D, the disturbance step size of the next time slot is equal to the product of the direction of the disturbance step and the current step threshold step, and the optimal phase of the transmitter in the next time slot is the optimal phase of the current time slot plus the disturbance step of the next time slot. Long and subtracting the perturbation step size of the current time slot, the transmitter transmits the signal by using the best phase of the time slot below the next time slot plus the initial phase as the transmission phase; wherein the direction of the disturbance step is randomly generated by the transmitter;

In step E, the disturbance step size of the next time slot is equal to the product of the optimal direction of the disturbance step and the current phase threshold step, and the optimal phase of the transmitter in the next time slot is the optimal phase of the current time slot plus the next time slot. Disturb the step size and subtract the perturbation step size of the current time slot. The transmitter transmits the signal at the optimum phase of a time slot below the next time slot plus the initial phase as the transmit phase. A second technical problem to be solved by the present invention is to provide a distributed system including a plurality of transmitters and receivers. The receiver is configured to determine, at each time slot, whether the strength of the received signal in the current time slot is greater than or equal to the next time. a threshold value, and feedback corresponding single-bit information to the transmitter according to the judgment result; the transmitter is configured to perform parsing after receiving the single-bit information;

If the optimal direction of the disturbance step is not determined, and the result of the analysis is that the strength of the received signal of the receiver is greater than or equal to the threshold of the next stage, the optimal phase of the transmitter in the next time slot still takes the best phase in the current time slot, and No disturbance step is generated in the next time slot, and the optimum phase of the next time slot is added to the next time slot plus the initial phase as the transmission phase to transmit the signal, and the number of stages is increased by one; the initial phase is the transmitter in the first The time slot is estimated based on the signal received by the transmitter; otherwise, the perturbation step size of the transmitter in the next time slot is equal to the product of its disturbance step direction and the current phase threshold step size, and the best phase of the transmitter in the next time slot is current. The optimal phase of the time slot is added to the disturbance step of the next time slot and the disturbance step size of the current time slot is subtracted, and the transmitter transmits the signal by adding the initial phase as the transmission phase to the optimum phase of the time slot below the next time slot; The direction of the disturbance step is randomly generated by the transmitter; the optimal direction of the disturbance step is when the received signal strength is less than the threshold of the next stage. Disturbance step after the optimum direction after all slots situation, when first received signal strength is greater than or equal to the threshold value the next stage, the direction of the disturbance step is more than one time slot;

If the optimal direction of the disturbance step is determined and the result of the analysis is that the strength of the received signal of the receiver is greater than or equal to the threshold of the next stage, the optimal phase of the transmitter in the next time slot still takes the best phase in the current time slot, and A time slot does not generate a disturbance step, and the signal is transmitted in the next time slot directly from the optimal phase of the next time slot plus the initial phase as the transmission phase, and the number of stages is incremented by one; the initial phase is when the transmitter is in the first time. The slot is estimated based on the signal received by the transmitter; otherwise, the perturbation step size of the transmitter in the next slot is equal to the product of the optimal direction of its perturbation step and the current phase threshold step, and the best phase of the transmitter in the next slot is The optimum phase of the current time slot is added to the disturbance step size of the next time slot and the disturbance step size of the current time slot is subtracted, and the transmitter transmits the signal by adding the initial phase as the transmission phase to the optimum phase of the time slot below the next time slot. Further, the stage threshold step size decreases as the number of stages increases.

Further, the correspondence between the number of stages, the stage threshold step, and the stage threshold is as follows:

Number of stages k (k ≤ S)

1

2

3

4

5

6

7

8

9

Stage threshold step g(k)	π/2	π/4	π/2 3	π/2 4	π/2 5	π/2 6	π/2 7	π/2 8	π/2 9
										Stage threshold Th(k)	0	1.4142	1.8478	1.9616	1.9904	1.9976	1.9994	1.9998	2.0000

Compared with the prior art, the invention has the following advantages:

1) For a given small step disturbance of the prior art, the present invention proposes a step size disturbance selection mechanism: a stage threshold step selection mechanism. In the initial few time slots, large-step perturbations can be used to increase the received signal strength of the receiver. In the late time slots, small step disturbances can be used to obtain better stability even if the receiver signal strength converges to the ideal condition.

2) The direction of the original technique to the transmitter's disturbance step is random, and both the direction and the size are adjusted. The present invention first determines the direction of the transmitter's disturbance step, and when the direction is determined, the adjustment of each transmitter is explained. The direction is correct, and only the disturbance step size is adjusted in the subsequent time slot, which can further accelerate the convergence of the subsequent received signal strength.

DRAWINGS

1 is an architectural diagram of a distributed system provided by the present invention;

2 is a chart of a step threshold selection mechanism provided by the present invention;

3 is a flow chart of feedback adjustment of a phase synchronization method provided by the present invention;

4 is a flow chart showing that RSS is greater than a current threshold in the phase synchronization method provided by the present invention;

FIG. 5 is a flowchart of the RSS synchronization method in the phase synchronization method provided by the present invention being smaller than the current threshold.

detailed description

The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It is understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

In response to a given small step size of the prior art, the present invention proposes a step selection mechanism that employs different step sizes at different stages to accelerate convergence at different received signal strengths. And, The original technology compensates the phase compensation direction of the transmitter randomly, and adjusts the direction and the size. The present invention proposes a method for determining the direction of the transmitter compensation phase, and the direction determination can accelerate the convergence of the subsequent received signal strength. .

1 shows the architecture of a distributed system to which the present invention is applied. A plurality of transmitters simultaneously transmit signals to a receiver, and adjust the phase of the transmitted signal of the next time slot according to the single bit information fed back by the receiver, and finally pass the phase. The manner of compensation synchronizes the phase of the respective received signals arriving at the receiver.

Specifically, in the single-bit feedback closed-loop system of FIG. 1, in the nth time slot, the transmitter receives the fed back single-bit information to determine whether to adjust the transmit signal phase. The transmit phase formula for each transmitter is as follows:

Φ _i =γ _i +θ _i (n)+ψ _i γ _i ,ψ _i ∈[0,2π) (1)

Where γ _i represents the unknown phase offset of the ith transmitter, ψ _i represents the phase response between the ith transmitter and the receiver, assuming that γ _i and ψ _i are always static during the convergence of the algorithm, in accordance with The variable of [0, 2π) is evenly distributed and is unknown to both the transmitter and the receiver, both of which are collectively referred to as the initial phase C. We will adjust the variable θ _i (n) of the nth time slot of the transmitter i by the single-bit feedback information from the receiver as the optimum phase. The initial value is set to θ _i (0)=0. Since the goal of the algorithm is to achieve phase synchronization at the receiver, assuming that each transmitter has unity power, the channel gain per transmitter and receiver is one. Therefore, the received signal strength of the receiver at time slot n is expressed as follows:

Where j denotes the root of -1, the best phase θ _i (n) is recorded in the memory, and at each time slot, a disturbance step δ _i (n) is compensated according to the single-bit information fed back by the receiver, ie The optimal phase θ _i (n) is a function of δ _i (n), enabling the algorithm to achieve phase synchronization at the receiver end in a certain time slot.

The present invention introduces a new disturbance step generation method to achieve phase compensation for each transmitter, that is, the disturbance step size δ _i (n) is staged and large to small. In the phase threshold step selection mechanism, the disturbance step size δ _i (n) is generated by the phase threshold step g(k) in FIG. 2, and the direction is random. During the training of the entire algorithm, if the received signal strength RSS of the current phase k is greater than the set phase threshold Th(k+1) (Table 1), then the training process proceeds to the next phase k+1. Explain that the current adjustments are useful to bring the phases between the transmitters closer together in a common direction.

First, the concept of phase threshold step and phase threshold should be distinguished. The phase threshold step is the size of the random disturbance added by the transmitter at different stages. The phase threshold is determined by the maximum phase difference x(k) between the transmitters by the formula (4). The value is at the receiver and the received signal. The intensity R(n) is compared to determine if it should proceed to the next stage. The two are mainly generated by the following two formulas:

x(k)=π/2 ^k k=0:1:S-1 (3)

Where S is the total number of phases in a given training process, N is the number of transmitters, and x(k) is the maximum phase difference between the transmitters in the kth phase. Correspondingly, Th(k) is in the first The minimum received signal strength RSS of the k-phase, which is the concept of the phase threshold. According to the maximum phase difference x(k), the phase threshold step g(k) of each phase can be generated, and the generation rule is:

g(k)=x(k)/2 (5)

That is, the phase threshold step is half of the maximum phase difference. It can be seen that in the case where the transmitter phase variation range is [0, 2π), the transmitter phase difference is [0, π). Obviously, the phase threshold step is half of the maximum phase difference for two purposes: It is possible to regularly reduce the phase difference between the transmitters. Second, the step size of the adjustment will be from large to small, the acceleration algorithm converges in the early stage, and the algorithm converges to the ideal situation in the later stage.

Figures 3, 4, and 5 illustrate the feedback adjustment mechanism between the transmitter and receiver of the present invention. Referring to FIG. 3 to FIG. 5 together. The closed-loop phase synchronization method based on directed negative feedback of the distributed system provided by the present invention comprises the following steps:

Step A: The receiver determines whether the strength of the received signal in the current time slot is greater than or equal to the threshold of the next stage, and feeds back corresponding single-bit information to the transmitter according to the determination result;

Step B: The transmitter parses the single-bit information, and if the optimal direction of the disturbance step is not determined, and the analysis result is that the strength of the received signal of the receiver is greater than or equal to the threshold of the next stage, Step C is performed, otherwise step D is performed; if the optimal direction of the disturbance step is determined, and the result of the analysis is that the strength of the received signal of the receiver is greater than or equal to the threshold of the next stage, step C is performed, otherwise step E is performed; The longest optimal direction is the case where the received signal strength is greater than or equal to the next stage threshold when the received signal strength is less than the next stage threshold. The disturbance step direction of the previous time slot is the disturbance step of all subsequent time slots. Optimal direction

Step C, the best phase of the transmitter in the next time slot still takes its best phase in the current time slot, and does not generate a disturbance step in the next time slot, and directly adds the initial phase to the best phase of the next time slot in the next time slot. Transmitting a signal as a phase of transmission, and adding one to the number of stages; the initial phase is obtained by estimating, based on a signal received by the transmitter, the transmitter in the first time slot;

In step D, the disturbance step size of the next time slot is equal to the product of the direction of the disturbance step and the current step threshold step, and the optimal phase of the transmitter in the next time slot is the optimal phase of the current time slot plus the disturbance step of the next time slot. Long and subtracting the perturbation step size of the current time slot, the transmitter transmits the signal by using the best phase of the time slot below the next time slot plus the initial phase as the transmission phase; wherein the direction of the disturbance step is randomly generated by the transmitter;

In step E, the disturbance step size of the next time slot is equal to the product of the optimal direction of the disturbance step and the current phase threshold step, and the optimal phase of the transmitter in the next time slot is the optimal phase of the current time slot plus the next time slot. Disturb the step size and subtract the perturbation step size of the current time slot. The transmitter transmits the signal at the optimum phase of a time slot below the next time slot plus the initial phase as the transmit phase. Steps C, D, and E correspond to the +1 module flow of FIG. 4 and the -1 module flow of FIG. 5, respectively. The invention first adjusts the direction of the disturbance step δ _i (n). After determining its direction, the direction of phase compensation of each transmitter will not change, but at different stages, according to the feedback information of the receiver of each time slot, Adjust the size of the disturbance step to accelerate the effective combination of the transmitter phase at the receiver.

In the directional negative feedback compensation closed-loop synchronization algorithm, we introduce two state variables state_a and state_b, and a direction variable Direct. Direct indicates the phase compensation direction of the transmitter in the time slot n of the disturbance step δ _i (n). We initialize the optimal phase θ _i (0) = 0, and the state variable states_a = 0 and state_b = 0 in the time slot n = 0.

When state_a=0, it indicates that the transmitter can smoothly jump to the next stage in each time slot from slot 0 to slot n, that is, R(n)≥Th(k+1). Physically, the receiver is feedback "+1" bit information from 0 time slot to n time slot, and no "-1" bit information is fed back during the period. When state_b=0, it indicates that the transmitter does not jump into the next stage for the first time, that is, the receiver feeds back "-1" bit information for the first time. If the subsequent time slot continuously feeds back "-1" bit information, state_b=0 remains unchanged. At the same time, the direction Direct of the step disturbance is also undetermined. There are two key changes in the training process of this algorithm. One is that the transmitter receives the "-1" bit information for the first time, and the other is from the time slot n to the time slot n+1. The first transition from "-1" to "+1" occurs, that is, the "n" receives "-1. "Feedback information, n+1 time to receive "+1" feedback information, and, for the first time in the entire process.

When the transmitter receives the "-1" feedback for the first time, state_a will be set to 1, ie state_a = 1, and will not change. During subsequent iterations, state_a will be used to control state_b to be set to 1 when it first appears from "-1" to "+1". State_b=1 indicates that the direction of the disturbance step is determined in a certain slot n, and at the same time, state_b will not change from the time of n+1, indicating that the subsequent training process will no longer need to adjust the direction of phase compensation of the transmitter. Each transmitter will adjust the disturbance step size in the determined direction.

As can be seen from the above, state_a controls the adjustment of state_b, and state_b controls the direction variable Direct. Direct adjusts the state of state_b only when the receiver feeds back negative bit information. When the transmitter receives the "-1" feedback, the Direct adjustment is as follows:

When state_b=0, we randomly generate the direction of the disturbance step, and when the transition from “-1” to “+1” occurs for the first time from time n to time n+1, Direct will remain unchanged and follow-up. The training process is no longer changed, ie the direction of the transmitter phase compensation has been determined.

Obviously, in addition to the adjustment of the above three variables, we need to adjust the other two variables throughout the training process: the perturbation step size adjusted according to the receiver feedback information, the direction variable Direct and the phase threshold step g(k). δ _i (n), the adjustment of the optimal phase θ _i (n). The adjustment formula for both is as follows:

δ _i (n+1)=Direct×g(k) R(n)<Th(k+1) (7)

When the transmitter receives the "-1" feedback, the step disturbance δ _i (n+1) at time n+1 is updated according to equations (5) and (6), ie δ _i (n+1)=Direct ×g(k). If the "+1" bit information is fed back, no disturbance step is generated. When the receiver feeds back "+1" information, according to formula (8), the optimum phase is kept constant, that is, θ _i (n+1) = θ _i (n). If the "-1" information is fed back, θ _i (n+1) = θ _i (n) + δ _i (n+1) - δ _i (n).

The present invention can be applied to a wireless environment in which a plurality of wireless transmitting nodes transmit the same signal to a receiver, form a closed loop by the feedback adjustment mechanism of the present invention, and quickly converge the signal strength of the receiver under a complicated channel.

The above is only the preferred embodiment of the present invention, and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection of the present invention. Within the scope.

Claims (6)

1. A closed-loop phase synchronization method based on directed negative feedback for a distributed system, comprising the following steps performed in each time slot:

   Step A: The receiver determines whether the strength of the received signal in the current time slot is greater than or equal to the threshold of the next stage, and feeds back corresponding single-bit information to the transmitter according to the determination result;

   Step B: After the transmitter receives the single-bit information, it performs parsing. If the optimal direction of the disturbance step is not determined, and the analysis result is that the strength of the received signal of the receiver is greater than or equal to the threshold of the next stage, step C is performed, otherwise execution is performed. Step D; if the optimal direction of the disturbance step is determined, and the analysis result is that the strength of the received signal of the receiver is greater than or equal to the threshold of the next stage, step C is performed, otherwise step E is performed; the optimal direction of the disturbance step is When the received signal strength is less than the next-stage threshold, when the first received signal strength is greater than or equal to the next-stage threshold, the disturbance step direction of the previous time slot is the optimal direction of the disturbance step of all subsequent time slots;

   Step C, the best phase of the transmitter in the next time slot still takes its best phase in the current time slot, and does not generate a disturbance step in the next time slot, and directly adds the initial phase to the best phase of the next time slot in the next time slot. Transmitting a signal as a phase of transmission, and adding one to the number of stages; the initial phase is obtained by estimating, based on a signal received by the transmitter, the transmitter in the first time slot;

   In step D, the disturbance step size of the next time slot is equal to the product of the direction of the disturbance step and the current step threshold step, and the optimal phase of the transmitter in the next time slot is the optimal phase of the current time slot plus the disturbance step of the next time slot. Long and subtracting the perturbation step size of the current time slot, the transmitter transmits the signal by using the best phase of the time slot below the next time slot plus the initial phase as the transmission phase; wherein the direction of the disturbance step is randomly generated by the transmitter;

   In step E, the disturbance step size of the next time slot is equal to the product of the optimal direction of the disturbance step and the current phase threshold step, and the optimal phase of the transmitter in the next time slot is the optimal phase of the current time slot plus the next time slot. Disturb the step size and subtract the perturbation step size of the current time slot. The transmitter transmits the signal at the optimum phase of a time slot below the next time slot plus the initial phase as the transmit phase.
2. The closed loop phase synchronization method of claim 1 wherein said phase threshold step size decreases as the number of stages increases.
3. The closed-loop phase synchronization method according to claim 2, wherein the correspondence between the number of stages, the stage threshold step, and the stage threshold is as follows:

   Number of stages k (k ≤ S) 1 2 3 4 5 6 7 8 9 Stage threshold step g(k) π/2 π/4 π/2 ³ π/2 ⁴ π/2 ⁵ π/2 ⁶ π/2 ⁷ π/2 ⁸ π/2 ⁹ Stage threshold Th(k) 0 1.4142 1.8478 1.9616 1.9904 1.9976 1.9994 1.9998 2.0000
4. A distributed system comprising a plurality of transmitters and receivers, wherein the receiver is configured to determine, in each time slot, whether the strength of the received signal in the current time slot is greater than or equal to a threshold value of the next stage, and according to the determination result Transmitting corresponding single bit information to the transmitter; the transmitter is configured to perform parsing after receiving the single bit information;

   If the optimal direction of the disturbance step is not determined, and the result of the analysis is that the strength of the received signal of the receiver is greater than or equal to the threshold of the next stage, the optimal phase of the transmitter in the next time slot still takes the best phase in the current time slot, and No disturbance step is generated in the next time slot, and the optimum phase of the next time slot is added to the next time slot plus the initial phase as the transmission phase to transmit the signal, and the number of stages is increased by one; the initial phase is the transmitter in the first The time slot is estimated based on the signal received by the transmitter; otherwise, the perturbation step size of the transmitter in the next time slot is equal to the product of its disturbance step direction and the current phase threshold step size, and the best phase of the transmitter in the next time slot is current. The optimal phase of the time slot is added to the disturbance step of the next time slot and the disturbance step size of the current time slot is subtracted, and the transmitter transmits the signal by adding the initial phase as the transmission phase to the optimum phase of the time slot below the next time slot; The direction of the disturbance step is randomly generated by the transmitter; the optimal direction of the disturbance step is when the received signal strength is less than the threshold of the next stage. After the condition, the first time when the received signal strength is greater than or equal to the threshold value the next stage, the disturbance step direction than a slot The optimal direction of the disturbance step for all subsequent time slots;

   If the optimal direction of the disturbance step is determined and the result of the analysis is that the strength of the received signal of the receiver is greater than or equal to the threshold of the next stage, the optimal phase of the transmitter in the next time slot still takes the best phase in the current time slot, and A time slot does not generate a disturbance step, and the signal is transmitted in the next time slot directly from the optimal phase of the next time slot plus the initial phase as the transmission phase, and the number of stages is incremented by one; the initial phase is when the transmitter is in the first time. The slot is estimated based on the signal received by the transmitter; otherwise, the perturbation step size of the transmitter in the next slot is equal to the product of the optimal direction of its perturbation step and the current phase threshold step, and the best phase of the transmitter in the next slot is The optimum phase of the current time slot is added to the disturbance step size of the next time slot and the disturbance step size of the current time slot is subtracted, and the transmitter transmits the signal by adding the initial phase as the transmission phase to the optimum phase of the time slot below the next time slot.
5. The distributed system of claim 4 wherein said phase threshold step size decreases as the number of stages increases.
6. The distributed system according to claim 5, wherein the correspondence between the number of stages, the stage threshold step, and the stage threshold is as follows:

   Number of stages k (k ≤ S) 1 2 3 4 5 6 7 8 9 Stage threshold step g(k) π/2 π/4 π/2 ³ π/2 ⁴ π/2 ⁵ π/2 ⁶ π/2 ⁷ π/2 ⁸ π/2 ⁹ Stage threshold Th(k) 0 1.4142 1.8478 1.9616 1.9904 1.9976 1.9994 1.9998 2.0000

Images