CN114337941A - A kind of channel resource allocation method in multi-packet combined transmission HARQ system - Google Patents

Abstract

The invention discloses a channel resource allocation method in a multi-packet combined transmission HARQ system, which comprises the following steps: modeling a transmission channel as an independent and identically distributed block fading channel, calculating the probability distribution of channel capacity, and modeling the probability distribution of errors between a channel capacity estimated value and an actual value as zero-mean normal distribution; before the nth time slot starts, a controller of the receiving end feeds back a data packet decoding result of the nth-1 time slot, accumulated mutual information and a channel capacity estimation value of the nth time slot to the transmitting end; in the nth time slot, the controller of the transmitting end calculates the allocation proportion of the channel resources according to the feedback information and the transmission turn of the data packet transmitted by the nth time slot; when 0< p <1, HOL and HOL-next packets share the timeslot channel resources in a time-shared manner and are transmitted in the same timeslot. The embodiment of the invention solves the problem that the existing channel resource allocation strategy is irrelevant to the real-time channel state and can not adapt to the change of the channel condition well.

Description

Translated fromChinese

一种多包合并传输HARQ系统中的信道资源分配方法A kind of channel resource allocation method in multi-packet combined transmission HARQ system

技术领域technical field

本发明涉及但不限于无线通信技术的资源分配控制技术领域，尤指一种多包合并传输HARQ系统中的信道资源分配方法。The present invention relates to, but is not limited to, the technical field of resource allocation control in wireless communication technology, and in particular, to a channel resource allocation method in a multi-packet combined transmission HARQ system.

背景技术Background technique

在传统混合自动重传请求(HARQ)模型中，每个时隙全部的符号资源都由一个数据包的子码字独享，然而有时发射端只需要分配少量的符号进行传输，接收端就可以获得足够的累计互信息，进而解码成功，这就产生了时隙资源的浪费。In the traditional hybrid automatic repeat request (HARQ) model, all the symbol resources of each time slot are exclusively shared by the sub-codewords of a data packet. However, sometimes the transmitter only needs to allocate a small number of symbols for transmission, and the receiver can Sufficient accumulated mutual information is obtained, and then decoding is successful, which results in a waste of time slot resources.

目前，时隙资源的浪费问题可以通过两种方式解决：一是通过收发双方协商动态地调整时隙长度，这种方案实现复杂并且会带来额外的信令开销；另一种方式是保持时隙长度恒定，多个用户共同使用每个时隙的信道资源，这种方案需要依赖上层的调度算法来协调不同用户间的时隙资源分配。在实际应用中，对于一块完整的资源空间，使用若干不定长的小块进行填充，最终实现完美填充是相当困难的。At present, the waste of time slot resources can be solved in two ways: one is to dynamically adjust the time slot length through negotiation between the sender and the receiver, which is complicated to implement and will bring additional signaling overhead; The slot length is constant, and multiple users jointly use the channel resources of each time slot. This scheme needs to rely on the scheduling algorithm of the upper layer to coordinate the allocation of time slot resources among different users. In practical applications, for a complete resource space, it is quite difficult to use several small blocks of indeterminate length for filling, and it is quite difficult to achieve perfect filling in the end.

发明内容SUMMARY OF THE INVENTION

本发明的目的：本发明实施例提出一种多包合并传输HARQ系统中的信道资源分配方法，以解决现有信道资源分配策略与实时信道状态无关，不能很好地适应信道条件的变化的问题。Purpose of the present invention: The embodiment of the present invention proposes a channel resource allocation method in a multi-packet combined transmission HARQ system, so as to solve the problem that the existing channel resource allocation strategy has nothing to do with the real-time channel state and cannot well adapt to changes in channel conditions .

本发明的技术方案：本发明实施例提出一种多包合并传输HARQ系统中的信道资源分配方法，包括：Technical solution of the present invention: The embodiment of the present invention proposes a channel resource allocation method in a multi-packet combined transmission HARQ system, including:

步骤A：将从发送端到接收端的传输信道建模为独立同分布的块衰落信道，并计算块衰落信道的信道容量c的概率分布，将信道容量估计值

与实际值间误差e的概率分布建模为零均值正态分布；其中，每个时隙的估计误差独立同分布；Step A: Model the transmission channel from the sender to the receiver as an independent and identically distributed block fading channel, and calculate the probability distribution of the channel capacity c of the block fading channel.

The probability distribution of the error e between the actual value and the actual value is modeled as a zero-mean normal distribution; wherein, the estimated error of each time slot is independent and identically distributed;

步骤B：在第n个时隙开始前，接收端的第二HARQ控制器将第n-1个时隙的数据包译码结果M[n-1]、累计互信息I[n-1]和对第n个时隙的信道容量估计值

反馈给发射端；Step B: Before the start of the nth time slot, the second HARQ controller at the receiving end decodes the data packet decoding result M[n-1] of the n-1th time slot, the accumulated mutual information I[n-1] and Channel capacity estimate for the nth slot

feedback to the transmitter;

步骤C：在第n个时隙，发射端的第一HARQ控制器根据步骤B中的反馈信息以及第n个时隙传输数据包的传输轮次，计算出第n个隙的信道资源分配比例p(0≤p≤1)；并将累计互信息、信道容量估计值c和决策值p离散化，通过信道容量值c和信道估计误差e的概率分布计算状态转移概率，采用无限阶段平均准则模型的值迭代算法求解最优的决策策略，获得离散状态与离散决策p的映射关系；Step C: In the nth time slot, the first HARQ controller at the transmitting end calculates the channel resource allocation ratio p of the nth slot according to the feedback information in Step B and the transmission round of the transmission data packet in the nth time slot. (0≤p≤1); discretize the accumulated mutual information, the estimated channel capacity c and the decision value p, calculate the state transition probability through the probability distribution of the channel capacity value c and the channel estimation error e, and use the infinite stage average criterion model The value iterative algorithm to solve the optimal decision-making strategy and obtain the mapping relationship between the discrete state and the discrete decision p;

步骤D：当0<p<1时，队首数据包(HOL数据包)和第二数据包(HOL-next数据包)以分时的方式共享时隙信道资源在同一时隙合并传输，当p＝0时，HOL数据包被丢弃，HOL-next数据包独享时隙信道资源传输，当p＝1时，HOL数据包独享信道资源完成重传。Step D: When 0<p<1, the first data packet (HOL data packet) and the second data packet (HOL-next data packet) share the time slot channel resources in a time-sharing manner and are combined and transmitted in the same time slot. When p=0, the HOL data packet is discarded, and the HOL-next data packet exclusively uses the time slot channel resource for transmission. When p=1, the HOL data packet exclusively uses the channel resource to complete the retransmission.

可选地，如上所述的一种多包合并传输HARQ系统中的信道资源分配方法中，Optionally, in the above-mentioned method for channel resource allocation in a multi-packet combined transmission HARQ system,

每个时隙传输数据包中包括：位于发送队列队首的队首数据包(HOL数据包)和位于发送队列中的第二数据包(HOL-next数据包)；其中，所述HOL数据包占据p·N_s个符号，所述HOL-next数据包占据(1-p)·N_s个符号。Each time slot transmission data packet includes: the head data packet (HOL data packet) located at the head of the sending queue and the second data packet (HOL-next data packet) located in the sending queue; wherein, the HOL data packet Occupying p·_Ns symbols, the HOL-next packet occupies (1-p)·_Ns symbols.

可选地，如上所述的一种多包合并传输HARQ系统中的信道资源分配方法中，所述步骤A包括：Optionally, in the above-mentioned method for channel resource allocation in a multi-packet combined transmission HARQ system, the step A includes:

步骤A1，将传输信道建模为独立同分布的瑞利衰落信道，则信道增益h服从瑞利分布，设定发射信号功率为ε_b，信道噪声功率为N₀，则接收端的信噪比γ表示为：Step A1, the transmission channel is modeled as an independent and identically distributed Rayleigh fading channel, then the channel gain h obeys the Rayleigh distribution, and the transmit signal power is set as ε_b , and the channel noise power is N₀ , then the signal-to-noise ratio at the receiving end is γ Expressed as:

其中，γ服从χ²分布，接收端的信噪比γ的概率密度函数表示为：Among them, γ obeys the χ² distribution, and the probability density function of the signal-to-noise ratio γ at the receiving end is expressed as:

其中，

表示接收端的平均信噪比；in,

Represents the average signal-to-noise ratio of the receiver;

步骤A2，根据香农公式，计算出接收端从每个符号可以获得的最大互信息为：Step A2, according to Shannon's formula, calculate the maximum mutual information that the receiver can obtain from each symbol as:

c＝log₂(1+γ)；c=log₂ (1+γ);

其中，c为信道容量，表示信息能够无差错传输的最大速率；通过复合随机变量的概率密度公式得出信道容量c的概率分布；Among them, c is the channel capacity, indicating the maximum rate at which information can be transmitted without errors; the probability distribution of the channel capacity c is obtained by the probability density formula of the composite random variable;

步骤A3，将信道容量估计值

与实际值间误差e的概率分布建模为零均值正态分布，并设定

与c的关系满足以下条件：Step A3, the estimated value of the channel capacity

The probability distribution of the error e from the actual value is modeled as a zero-mean normal distribution, and set

The relationship with c satisfies the following conditions:

其中，e表示信道容量的估计误差。Among them, e represents the estimation error of the channel capacity.

可选地，如上所述的一种多包合并传输HARQ系统中的信道资源分配方法中，所述步骤A中，Optionally, in the above-mentioned method for channel resource allocation in a multi-packet combined transmission HARQ system, in the step A,

设定不同时隙的信道容量的估计误差e互不相关，所有时隙的估计误差均服从相同的零均值正太分布，即

σ表示估计误差的标准差。It is assumed that the estimated errors e of the channel capacity of different time slots are independent of each other, and the estimated errors of all time slots obey the same zero-mean normal distribution, that is,

σ represents the standard deviation of the estimation error.

可选地，如上所述的一种多包合并传输HARQ系统中的信道资源分配方法中，数据包译码结果的确认信息有两种取值：ACK/NACK；所述步骤B包括：Optionally, in the above-mentioned method for channel resource allocation in a multi-packet combined transmission HARQ system, the acknowledgment information of the data packet decoding result has two values: ACK/NACK; the step B includes:

步骤B1，用u_l和u_l+1分别表示当前时隙的HOL数据包和HOL-next数据包，l表示数据包编号，用

表示数据包u_l在第k_l轮次传输后的确认信息；并根据第n-1个时隙的信道资源分配比例p反馈数据包译码结果的确认信息；Step B1, use u_l and u_l+1 to represent the HOL data packet and HOL-next data packet of the current time slot, respectively, and l represents the data packet number, and use

Represents the acknowledgment information of the data packet ul after the_k1th round of transmission; and feeds back the acknowledgment information of the decoding result of the data packet according to the channel resource allocation ratio p of the n-_1th time slot;

步骤B2，用

表示数据包u_l在第k_l轮次传输后的接收端所需的累计互信息，并根据重传机制计算反馈信息中的累计互信息；Step B2, use

represents the accumulated mutual information required by the receiver after the data packet u_l is transmitted in the k_lth round, and calculates the accumulated mutual information in the feedback information according to the retransmission mechanism;

步骤B3，接收端在第n个隙即将开始前，测得第n个时隙的信道容量估计

并将

与第n-1个时隙的数据包译码结果和累计互信息一并反馈给发射端。Step B3, the receiving end measures the channel capacity estimate of the nth time slot just before the nth time slot starts

and will

It is fed back to the transmitter together with the data packet decoding result of the n-1th time slot and the accumulated mutual information.

可选地，如上所述的一种多包合并传输HARQ系统中的信道资源分配方法中，所述步骤B1中，Optionally, in the above-mentioned method for channel resource allocation in a multi-packet combined transmission HARQ system, in step B1,

当0<p<1时，称为TS模式，接收端有两个数据包译码结果，即同时反馈两个数据包的确认信息

和

When 0<p<1, it is called TS mode, and the receiving end has two data packet decoding results, that is, the confirmation information of the two data packets is fed back at the same time.

and

当p＝1时，称为1P模式，只反馈数据包的确认信息

When p=1, it is called 1P mode, and only the acknowledgment information of the data packet is fed back

当p＝0时，称为0P模式，只反馈数据包的确认信息

When p=0, it is called 0P mode, and only the confirmation information of the data packet is fed back

用M[n]表示第n个时隙反馈的确认信息中的数据包译码结果，则其M[n]表示为：M[n] is used to represent the decoding result of the data packet in the acknowledgment information fed back by the nth time slot, then M[n] is represented as:

可选地，如上所述的一种多包合并传输HARQ系统中的信道资源分配方法中，所述步骤B2中，Optionally, in the above-mentioned method for channel resource allocation in a multi-packet combined transmission HARQ system, in step B2,

基于最多允许一次重传的情况，在TS模式下，HOL数据包必为重传数据包，HOL数据包在下个时隙必被丢弃，则反馈信息中的累计互信息I[n]表示为：Based on the fact that at most one retransmission is allowed, in TS mode, the HOL data packet must be a retransmission data packet, and the HOL data packet must be discarded in the next time slot, then the accumulated mutual information I[n] in the feedback information is expressed as:

第n个时隙的HOL数据包的累计互信息

与HOL-next数据包的累计互信息

通过前一时隙的各自的累计互信息递推得到，递推关系为：Cumulative Mutual Information of HOL Packets in the nth Slot

Cumulative mutual information with HOL-next packets

It is obtained by recursion of the respective accumulated mutual information of the previous time slot, and the recurrence relation is:

其中，n_l和n_l+1分别表示数据包u_l和数据包u_l+1初次传输的时隙编号。Wherein, n_l and n_l+1 respectively represent the time slot numbers of the first transmission of the data packet ul and the data packet ul_+1.

可选地，如上所述的一种多包合并传输HARQ系统中的信道资源分配方法中，所述的步骤C包括：Optionally, in the above-mentioned method for channel resource allocation in a multi-packet combined transmission HARQ system, the step C includes:

步骤C1，将步骤B中的反馈信息和传输轮次的值映射到相应的离散区间，得出当前的HARQ系统的离散状态；Step C1, the feedback information in step B and the value of the transmission round are mapped to the corresponding discrete interval, and the discrete state of the current HARQ system is obtained;

步骤C2，根据步骤C1所计算的离散状态到离散决策p的映射关系查找最优决策策略的p值。In step C2, the p value of the optimal decision-making strategy is searched according to the mapping relationship between the discrete state and the discrete decision p calculated in step C1.

可选地，如上所述的一种多包合并传输HARQ系统中的信道资源分配方法中，所述步骤D包括：Optionally, in the above-mentioned method for channel resource allocation in a multi-packet combined transmission HARQ system, the step D includes:

步骤D1，根据查找的最优决策策略的p值，确定当前的传输模式；Step D1: Determine the current transmission mode according to the p-value of the searched optimal decision-making strategy;

步骤D2，1P模式，整个传输时隙N_s个符号全部分配给HOL数据包的子码字；Step D2, 1P mode, the N_s symbols of the entire transmission time slot are all allocated to the subcodeword of the HOL data packet;

0P模式，整个传输时隙的符号资源全部分配给HOL-next数据包的子码字；0P mode, the symbol resources of the entire transmission time slot are all allocated to the sub-codewords of the HOL-next data packet;

TS模式，整个传输时隙的符号由HOL和HOL-next两个数据包的子码字共同享有，分别占据p·N_s个符号和(1-p)·N_s个符号，两个数据包的信息比特在编码时互不干扰，并且调制时的符号映射是互不重叠的。In TS mode, the symbols of the entire transmission time slot are shared by the sub-codewords of the two data packets HOL and HOL-next, occupying p·N_s symbols and (1-p)·N_s symbols respectively, two data packets The information bits do not interfere with each other during coding, and the symbol mappings during modulation do not overlap with each other.

本发明的有益效果：本发明实施例提出一种多包合并传输HARQ系统中的信道资源分配方法，该方法中引入了实时CSI估计反馈，系统根据确认信息，传输轮次、累计互信息和实时CSI估计四大参量，动态地调整传输模式和分配符号资源。为了获得最优的传输策略，将多包合并传输HARQ系统的传输过程建模为马尔可夫决策过程，利用值迭代算法解决了符号分配参数的优化问题。该方法对提高HARQ系统的吞吐率性能具有重要意义。Beneficial effects of the present invention: The embodiment of the present invention proposes a channel resource allocation method in a multi-packet combined transmission HARQ system, which introduces real-time CSI estimation feedback, and the system transmits rounds, accumulated mutual information and real-time CSI estimates four parameters, dynamically adjusts the transmission mode and allocates symbol resources. In order to obtain the optimal transmission strategy, the transmission process of the multi-packet combined transmission HARQ system is modeled as a Markov decision process, and the optimization problem of the symbol allocation parameters is solved by using the value iteration algorithm. This method is of great significance to improve the throughput performance of HARQ system.

与现有的基于累计互信息的技术相比，本发明实现了对CSI估计的利用，基于该指标对时隙的信道资源分配进行了实时的调节。在时隙信道资源固定的情况下，大幅提高了多包合并传输HARQ系统的信道资源利用率。本发明对提高HARQ系统的吞吐率性能，优化时隙信道资源分配具有重要意义，在无线通信系统中具有广阔的应用前景。Compared with the existing technology based on accumulating mutual information, the present invention realizes the utilization of CSI estimation, and adjusts the channel resource allocation of the time slot in real time based on the index. Under the condition that the time slot channel resources are fixed, the channel resource utilization rate of the multi-packet combined transmission HARQ system is greatly improved. The invention has important significance for improving the throughput performance of the HARQ system and optimizing the allocation of time slot channel resources, and has broad application prospects in the wireless communication system.

附图说明Description of drawings

附图用来提供对本发明技术方案的进一步理解，并且构成说明书的一部分，与本申请的实施例一起用于解释本发明的技术方案，并不构成对本发明技术方案的限制。The accompanying drawings are used to provide a further understanding of the technical solutions of the present invention, and constitute a part of the specification. They are used to explain the technical solutions of the present invention together with the embodiments of the present application, and do not limit the technical solutions of the present invention.

图1为本发明实施例提供的多包合并传输HARQ系统中的信道资源分配方法的流程图；1 is a flowchart of a channel resource allocation method in a multi-packet combined transmission HARQ system provided by an embodiment of the present invention;

图2为本发明实施例提供的基于CSI估计方法和基于AMI方法的系统吞吐率性能随平均信噪比变化的曲线图。FIG. 2 is a graph showing the variation of the system throughput performance with the average signal-to-noise ratio based on the CSI estimation method and the AMI-based method according to an embodiment of the present invention.

具体实施方式Detailed ways

为使本发明的目的、技术方案和优点更加清楚明白，下文中将结合附图对本发明的实施例进行详细说明。需要说明的是，在不冲突的情况下，本申请中的实施例及实施例中的特征可以相互任意组合。In order to make the objectives, technical solutions and advantages of the present invention clearer, the embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be noted that, the embodiments in the present application and the features in the embodiments may be arbitrarily combined with each other if there is no conflict.

在附图的流程图示出的步骤可以在诸如一组计算机可执行指令的计算机系统中执行。并且，虽然在流程图中示出了逻辑顺序，但是在某些情况下，可以以不同于此处的顺序执行所示出或描述的步骤。The steps shown in the flowcharts of the figures may be performed in a computer system, such as a set of computer-executable instructions. Also, although a logical order is shown in the flowcharts, in some cases the steps shown or described may be performed in an order different from that herein.

上述背景技术中已经说明现有解决时隙资源的浪费问题的技术方案所存在的普遍问题。另外，对于一块完整的资源空间，使用若干不定长的小块进行填充，最终实现完美填充是相当困难的。The general problems existing in the existing technical solutions for solving the waste of time slot resources have been described in the above background art. In addition, for a complete resource space, it is quite difficult to use several small blocks of indeterminate length for filling, and it is quite difficult to achieve perfect filling in the end.

针对上述问题，本发明实施例中考虑采用包合并传输HARQ技术解决上述时隙资源的浪费问题，该包合并传输HARQ技术中，每个时隙允许发射机同时处理多个数据包的子码字，将多个数据包的编码比特序列合并在同一个时隙传输。In view of the above problems, in the embodiment of the present invention, the HARQ technology for combined packet transmission is considered to solve the above-mentioned waste of time slot resources. In the HARQ technology for combined packet transmission, each time slot allows the transmitter to process the subcodewords of multiple data packets at the same time. , the coded bit sequences of multiple data packets are combined in the same time slot for transmission.

多包合并传输HARQ技术的一个核心问题就是合并传输时的信道资源分配算法，其通过调整同一时隙的传输信号中多个数据包的符号占比，以获得最优的系统性能。Jabi M于2015年在IEEE Transactions on Communications期刊发表了“Multipacket hybridARQ:Closing gap to the ergodic capacity”，提出了一种多包合并传输HARQ技术中基于AMI的信道资源分配算法，其主要思想为：一个数据包在接收端的累计互信息越高，说明该数据包距离译码成功所需的互信息越少，因此重传所需的互信息越少，相应的可以分配较少的符号资源，反之则反。但是这种信道资源分配策略与实时信道状态无关，不能很好地适应信道条件的变化。为了实现这一点，本发明实施例提供的方法在信道资源分配策略的控制参量中引入CSI估计。One of the core problems of the HARQ technology for multi-packet combined transmission is the channel resource allocation algorithm during combined transmission, which achieves optimal system performance by adjusting the symbol ratio of multiple data packets in the transmission signal of the same time slot. Jabi M published "Multipacket hybridARQ: Closing gap to the ergodic capacity" in IEEE Transactions on Communications in 2015, and proposed a channel resource allocation algorithm based on AMI in the HARQ technology of multi-packet combined transmission. The main idea is: a The higher the accumulated mutual information of the data packet at the receiving end, the less the mutual information required for the successful decoding of the data packet, the less the mutual information required for retransmission, and the less symbol resources can be allocated accordingly, and vice versa. opposite. However, this channel resource allocation strategy has nothing to do with the real-time channel state, and cannot well adapt to changes in channel conditions. In order to achieve this, the method provided by the embodiment of the present invention introduces CSI estimation into the control parameter of the channel resource allocation strategy.

本发明提供以下几个具体的实施例可以相互结合，对于相同或相似的概念或过程可能在某些实施例不再赘述。The present invention provides the following specific embodiments that can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments.

图1为本发明实施例提供的多包合并传输HARQ系统中的信道资源分配方法的流程图。本发明实施例提供的多包合并传输HARQ系统中的信道资源分配方法，主要包括：包括初始基值离散化处理、随机排序和序列映射等步骤，具体包括如下步骤：FIG. 1 is a flowchart of a channel resource allocation method in a multi-packet combined transmission HARQ system according to an embodiment of the present invention. The channel resource allocation method in the multi-packet combined transmission HARQ system provided by the embodiment of the present invention mainly includes: including initial base value discretization processing, random sorting, sequence mapping and other steps, and specifically includes the following steps:

步骤A：将从发送端到接收端的传输信道建模为独立同分布的块衰落信道，该块衰落信道的信道增益满足瑞利分布；另外，根据香农定理计算块衰落信道的信道容量c的概率分布，将信道容量估计值

与实际值间误差e的概率分布建模为零均值正态分布；其中，每个时隙的估计误差独立同分布；Step A: Model the transmission channel from the sender to the receiver as an independent and identically distributed block fading channel, and the channel gain of the block fading channel satisfies the Rayleigh distribution; in addition, calculate the probability of the channel capacity c of the block fading channel according to Shannon's theorem distribution, the channel capacity estimate

The probability distribution of the error e between the actual value and the actual value is modeled as a zero-mean normal distribution; wherein, the estimated error of each time slot is independent and identically distributed;

步骤B：在第n个时隙开始前，接收端的第二HARQ控制器将第n-1个时隙的数据包译码结果M[n-1]、累计互信息I[n-1]和对第n个时隙的信道容量估计值

反馈给发射端；Step B: Before the start of the nth time slot, the second HARQ controller at the receiving end decodes the data packet decoding result M[n-1] of the n-1th time slot, the accumulated mutual information I[n-1] and Channel capacity estimate for the nth slot

feedback to the transmitter;

步骤C：在第n个时隙，发射端的第一HARQ控制器根据步骤B中的反馈信息以及第n个时隙传输数据包的传输轮次，计算出第n个隙的信道资源分配比例p(0≤p≤1)；其中，位于发送队列队首的队首数据包(HOL数据包)占据p·N_s个符号，位于发送队列中的第二数据包(HOL-next数据包)占据(1-p)·N_s个符号。Step C: In the nth time slot, the first HARQ controller at the transmitting end calculates the channel resource allocation ratio p of the nth slot according to the feedback information in Step B and the transmission round of the transmission data packet in the nth time slot. (0≤p≤1); wherein, the head data packet (HOL data packet) located at the head of the sending queue occupies p·N_s symbols, and the second data packet (HOL-next data packet) located in the sending queue occupies (1-p)·N_s symbols.

该步骤C中，还将累计互信息、信道容量估计值c和决策值p离散化，通过信道容量值c和信道估计误差e的概率分布计算状态转移概率，采用无限阶段平均准则模型的值迭代算法求解最优的决策策略，获得离散状态与离散决策p的映射关系。In this step C, the accumulated mutual information, the channel capacity estimation value c and the decision value p are also discretized, the state transition probability is calculated through the probability distribution of the channel capacity value c and the channel estimation error e, and the value iteration of the infinite stage average criterion model is adopted. The algorithm solves the optimal decision-making strategy and obtains the mapping relationship between the discrete state and the discrete decision p.

步骤D：当0<p<1时，队首数据包(HOL数据包)和第二数据包(HOL-next数据包)以分时的方式共享时隙信道资源在同一时隙合并传输，当p＝0时，HOL数据包被丢弃，HOL-next数据包独享时隙信道资源传输，当p＝1时，HOL数据包独享信道资源完成重传。Step D: When 0<p<1, the first data packet (HOL data packet) and the second data packet (HOL-next data packet) share the time slot channel resources in a time-sharing manner and are combined and transmitted in the same time slot. When p=0, the HOL data packet is discarded, and the HOL-next data packet exclusively uses the time slot channel resource for transmission. When p=1, the HOL data packet exclusively uses the channel resource to complete the retransmission.

在本发明实施例中，上述步骤A的具体实施方式，可以包括：In this embodiment of the present invention, the specific implementation of the above step A may include:

步骤A1，将传输信道建模为独立同分布的瑞利衰落信道，则信道增益h服从瑞利分布，设定发射信号功率为ε_b，信道噪声功率为N₀，则接收端的信噪比γ表示为：Step A1, the transmission channel is modeled as an independent and identically distributed Rayleigh fading channel, then the channel gain h obeys the Rayleigh distribution, and the transmit signal power is set as ε_b , and the channel noise power is N₀ , then the signal-to-noise ratio at the receiving end is γ Expressed as:

因此，γ服从χ²分布，接收端的信噪比γ的概率密度函数表示为：Therefore, γ obeys the χ² distribution, and the probability density function of the signal-to-noise ratio γ at the receiving end is expressed as:

其中，

表示接收端的平均信噪比；in,

Represents the average signal-to-noise ratio of the receiver;

步骤A2，根据香农公式，计算出接收端从每个符号可以获得的最大互信息为：Step A2, according to Shannon's formula, calculate the maximum mutual information that the receiver can obtain from each symbol as:

c＝log₂(1+γ)；c=log₂ (1+γ);

其中，c为信道容量，表示信息能够无差错传输的最大速率；通过复合随机变量的概率密度公式得出信道容量c的概率分布；Among them, c is the channel capacity, indicating the maximum rate at which information can be transmitted without errors; the probability distribution of the channel capacity c is obtained by the probability density formula of the composite random variable;

步骤A3，将信道容量估计值

与实际值间误差e的概率分布建模为零均值正态分布，并设定

与c的关系满足以下条件：Step A3, the estimated value of the channel capacity

The probability distribution of the error e from the actual value is modeled as a zero-mean normal distribution, and set

The relationship with c satisfies the following conditions:

其中，e表示信道容量的估计误差。Among them, e represents the estimation error of the channel capacity.

需要说明的是，为了简化分析，设定不同时隙的信道容量的估计误差e互不相关，所有时隙的估计误差均服从相同的零均值正太分布，即

σ表示估计误差的标准差。It should be noted that, in order to simplify the analysis, it is assumed that the estimation errors e of the channel capacities of different time slots are independent of each other, and the estimation errors of all time slots obey the same zero-mean normal distribution, that is,

σ represents the standard deviation of the estimation error.

在本发明实施例中，上述步骤B中的数据包译码结果的确认信息有两种取值，即：ACK/NACK；上述步骤B的具体实施方式，可以包括：In the embodiment of the present invention, the confirmation information of the data packet decoding result in the above step B has two values, namely: ACK/NACK; the specific implementation manner of the above step B may include:

步骤B1，用u_l和u_l+1分别表示当前时隙的HOL数据包和HOL-next数据包，l表示数据包编号，用

表示数据包u_l在第k_l轮次传输后的确认信息；并根据第n-1个时隙的信道资源分配比例p反馈数据包译码结果的确认信息。Step B1, use u_l and u_l+1 to represent the HOL data packet and HOL-next data packet of the current time slot, respectively, and l represents the data packet number, and use

Indicates the acknowledgment information of the data packet ul after the_k1th round of transmission; and feeds back the acknowledgment information of the decoding result of the data packet according to the channel resource allocation ratio p of the n-₁ th time slot.

该步骤B1中，当0<p<1时，称为TS模式，接收端有两个数据包译码结果，即同时反馈两个数据包的确认信息

和

当p＝1时，称为1P模式，只反馈数据包的确认信息

当p＝0时，称为0P模式，只反馈数据包的确认信息

因此，用M[n]表示第n个时隙反馈的确认信息中的数据包译码结果，则其M[n]表示为：In this step B1, when 0<p<1, it is called TS mode, and the receiving end has two data packet decoding results, that is, the confirmation information of the two data packets is fed back at the same time.

and

When p=1, it is called 1P mode, and only the acknowledgment information of the data packet is fed back

When p=0, it is called 0P mode, and only the confirmation information of the data packet is fed back

Therefore, M[n] is used to represent the decoding result of the data packet in the acknowledgment information fed back by the nth time slot, then M[n] is represented as:

步骤B2，用

表示数据包u_l在第k_l轮次传输后的接收端所需的累计互信息，并根据重传机制计算反馈信息中的累计互信息。Step B2, use

Represents the accumulated mutual information required by the receiver after the data packet u_l is transmitted in the k_lth round, and calculates the accumulated mutual information in the feedback information according to the retransmission mechanism.

该步骤B2中，考虑到最多允许一次重传的情况，在TS模式下，HOL数据包必为重传数据包，由于其在下个时隙必被丢弃，因而对于发射机而言知晓其累计互信息并无意义。所以反馈信息中的累计互信息I[n]表示为：In this step B2, considering that at most one retransmission is allowed, in the TS mode, the HOL data packet must be a retransmission data packet. Since it must be discarded in the next time slot, the transmitter knows its cumulative mutual Information is meaningless. Therefore, the accumulated mutual information I[n] in the feedback information is expressed as:

第n个时隙的HOL数据包的累计互信息

与HOL-next数据包的累计互信息

通过前一时隙(即第n-1个)的各自的累计互信息递推得到，递推关系为：Cumulative Mutual Information of HOL Packets in the nth Slot

Cumulative mutual information with HOL-next packets

It is obtained by recursion of the respective accumulated mutual information of the previous time slot (that is, the n-1th), and the recurrence relation is:

其中，n_l和n_l+1分别表示数据包u_l和数据包u_l+1初次传输的时隙编号。显然I[n]的计算依赖于本时隙的信道容量c[n]，在第n个时隙c[n]在接收端是可以准确测得的，而初始累计互信息I_l,1＝I_l+1,1＝0，p和N_s是已知的系统参数，所以在每个时隙，对于接收端

和

也是已知的。Wherein, n_l and n_l+1 respectively represent the time slot numbers of the first transmission of the data packet ul and the data packet ul_+1. Obviously, the calculation of I[n] depends on the channel capacity c[n] of the current time slot. In the nth time slot c[n] can be accurately measured at the receiving end, and the initial accumulated mutual information I_l,1 = I_l+1,1 = 0, p and N_s are known system parameters, so in each time slot, for the receiver

and

is also known.

步骤B3，接收端在第n个隙即将开始前，测得第n个时隙的信道容量估计

并将

与第n-1个时隙的数据包译码结果和累计互信息一并反馈给发射端。Step B3, the receiving end measures the channel capacity estimate of the nth time slot just before the nth time slot starts

and will

It is fed back to the transmitter together with the data packet decoding result of the n-1th time slot and the accumulated mutual information.

在本发明实施例中，上述步骤C的具体实施方式，可以包括：In this embodiment of the present invention, the specific implementation of the above step C may include:

步骤C1，将步骤B中的反馈信息和传输轮次的值映射到相应的离散区间，得出当前的HARQ系统的离散状态；Step C1, the feedback information in step B and the value of the transmission round are mapped to the corresponding discrete interval, and the discrete state of the current HARQ system is obtained;

步骤C2，根据步骤C1所计算的离散状态到离散决策p的映射关系查找最优决策策略的p值。In step C2, the p value of the optimal decision-making strategy is searched according to the mapping relationship between the discrete state and the discrete decision p calculated in step C1.

本发明实施例在实际应用中，所述步骤C1包括以下6个子步骤：In practical applications of the embodiment of the present invention, the step C1 includes the following 6 sub-steps:

步骤C1-1：定义马尔可夫决策过程的状态空间，译码结果的取值集合M表示为：Step C1-1: Define the state space of the Markov decision process, and the value set M of the decoding result is expressed as:

M＝{NACK,ACK,(NACK,NACK),(NACK,ACK),(ACK,NACK),(ACK,ACK)}；M={NACK,ACK,(NACK,NACK),(NACK,ACK),(ACK,NACK),(ACK,ACK)};

用k₁和k₂分别表示HOL数据包和HOL-next数据包的传输轮次，基于最大传输次数K＝2，不允许两个数据包同时开启首次传输的情况，向量(k₁,k₂)的取值空间为：Use k₁ and k₂ to represent the transmission rounds of HOL data packets and HOL-next data packets, respectively. Based on the maximum transmission times K=2, it is not allowed to start the first transmission of two data packets at the same time. The vector (k₁ , k₂ ) value space is:

K＝＝{(1,0),(2,0),(2,1)}；K=={(1,0),(2,0),(2,1)};

将累计互信息的取值空间I离散化为J+1个离散值

其中

其余值满足：Discretize the value space I of accumulated mutual information into J+1 discrete values

in

The remaining values satisfy:

当接收端累计互信息为

时，确认信息为ACK，对于一个译码成功的数据包，反馈其累计互信息没有意义，在此类状态中可以不记录累计互信息参量。因此可以认为累计互信息的离散空间为

When the accumulated mutual information at the receiver is

When the acknowledgment information is ACK, for a successfully decoded data packet, it is meaningless to feed back its accumulated mutual information. In such a state, the accumulated mutual information parameter may not be recorded. Therefore, it can be considered that the discrete space of accumulated mutual information is

尽管信道容量的取值上限为正无穷，但是超过一定大小，其概率极小，因此可以设定一个有限的上限c_max，使其满足：Although the upper limit of the channel capacity is positive infinity, the probability of exceeding a certain size is extremely small, so a finite upper limit c_max can be set to satisfy:

Pr(c>c_max)<ε；Pr(c>c_max )<ε;

其中ε为一个趋于0的任意小正实数，然后在(0,c_max]区间内将

离散化为L个离散值

则信道容量估计的离散空间

where ε is an arbitrary small positive real number tending to 0, and then in the interval (0, c_max ], the

Discretize into L discrete values

Then the discrete space for channel capacity estimation

系统的状态信息是上述四个信息的组合，因此系统的离散状态空间S是这四个信息离散取值空间的笛卡尔积：The state information of the system is the combination of the above four information, so the discrete state space S of the system is the Cartesian product of the discrete value spaces of these four information:

步骤C1-2：对决策空间进行离散化处理。根据p的取值可将决策空间A分为三部分，

和

Step C1-2: Discretize the decision space. According to the value of p, the decision space A can be divided into three parts,

and

步骤C1-3：当接收端累计互信息超过数据包的比特大小N_b时代表该数据包可以译码成果，该马尔可夫决策过程的状态转移概率最终可以通过下式计算。Step C1-3: When the accumulated mutual information of the receiving end exceeds the bit size N_b of the data packet, it means that the data packet can be decoded, and the state transition probability of the Markov decision process can be finally calculated by the following formula.

其中等式所表示的事件可以通过下式近似计算：where the event represented by the equation can be approximated by:

其中

和Δ_I分别为离散空间C_d和I_d的离散粒度，系统状态转移概率可以通过随机变量e和随机变量

的概率分布求得。in

and ΔI are the discrete granularity of discrete space C_d and_Id , respectively, and the state transition probability of the system can be determined by the random variable e and the random variable

The probability distribution of .

步骤C1-4：定义报酬函数。系统的报酬r(s,s′,a)定义从状态s转移到s′的过程中每符号承载的成功解码信息量。这个报酬实际上仅与目标状态s′相关，当s′∈S_{NACK,ACK,(2,1)}∪S_ACk,(1,0)∪S_ACk,(2,0)时，都是仅有一个数据包传输成功，因此报酬为R；当s′∈S_{ACK,ACK,(2,1)}时，两个数据包同时传输成功，因此报酬为2R。而其他情况下没有数据包成功解码，因此报酬为0。报酬函数可归纳如下：Step C1-4: Define the reward function. The reward r(s, s', a) of the system defines the amount of successfully decoded information carried per symbol during the transition from state s to s'. This reward is actually only related to the target state s', when s'∈S_{NACK,ACK,(2,1)∪S ACk,(1,0)∪S ACk,(2,0)} , it is only One data packet is successfully transmitted, so the reward is R; when s′∈S_{ACK,ACK,(2,1)} , two data packets are transmitted successfully at the same time, so the reward is 2R. In other cases, no packets are successfully decoded, so the reward is 0. The reward function can be summarized as follows:

当前状态为s，采取行动a所能获得的平均报酬可由下式计算：The current state is s, and the average reward that can be obtained by taking action a can be calculated by the following formula:

步骤C1-5：定义准则函数。评估HARQ系统的吞吐率性能适用于无限阶段平均准则模型。系统的长期平均吞吐率定义如下：Step C1-5: Define the criterion function. Evaluating the throughput performance of a HARQ system applies an infinite-stage averaging criterion model. The long-term average throughput of the system is defined as follows:

其中n为时隙编号，r(s[n],π(s[n]))表示在时隙n采取决策π(s[n])获得的平均报酬。where n is the slot number, and r(s[n], π(s[n])) represents the average reward for taking the decision π(s[n]) at slot n.

步骤C1-6：根据前五个步骤的建模信息，使用马尔可夫决策过程无限阶段平均准则模型的值迭代算法求解最优策略。Step C1-6: According to the modeling information of the first five steps, use the value iteration algorithm of the Markov decision process infinite stage average criterion model to solve the optimal policy.

在本发明实施例中，上述步骤D的具体实施方式，可以包括：In this embodiment of the present invention, the specific implementation of the above step D may include:

步骤D1，根据查找的最优决策策略的p值，确定当前的传输模式；若p＝1，此时传输过程与传统HARQ无异，整个传输时隙的符号被HOL数据包的子码字完全占据，称之为1P模式；In step D1, the current transmission mode is determined according to the p value of the searched optimal decision-making strategy; if p=1, the transmission process is no different from the traditional HARQ at this time, and the symbols of the entire transmission time slot are completely replaced by the subcode words of the HOL data packet. Occupy, called 1P mode;

步骤D2，1P模式，整个传输时隙N_s个符号全部分配给HOL数据包的子码字；Step D2, 1P mode, the N_s symbols of the entire transmission time slot are all allocated to the subcodeword of the HOL data packet;

0P模式，整个传输时隙的符号资源全部分配给HOL-next数据包的子码字，意味着HoL数据包被丢弃了。当然在实际系统中，HoL数据包可能会由上层协议重新推入发送队列，等待之后的时隙完成传输；In 0P mode, the symbol resources of the entire transmission time slot are all allocated to the sub-codewords of the HOL-next data packet, which means that the HoL data packet is discarded. Of course, in the actual system, the HoL data packet may be re-pushed into the sending queue by the upper-layer protocol, waiting for the subsequent time slot to complete the transmission;

TS模式，整个传输时隙的符号由HOL和HOL-next两个数据包的子码字共同享有，分别占据p·N_s个符号和(1-p)·N_s个符号，两个数据包的信息比特在编码时互不干扰，并且调制时的符号映射是互不重叠的，这意味着两个数据包在接收端解调译码时可以独立完成，互不影响。In TS mode, the symbols of the entire transmission time slot are shared by the sub-codewords of the two data packets HOL and HOL-next, occupying p·N_s symbols and (1-p)·N_s symbols respectively, two data packets The information bits do not interfere with each other during encoding, and the symbol mapping during modulation does not overlap with each other, which means that the two data packets can be completed independently during demodulation and decoding at the receiving end without affecting each other.

本发明实施例提供一种多包合并传输HARQ系统中的信道资源分配方法，该方法中引入了实时CSI估计反馈，系统根据确认信息，传输轮次、累计互信息和实时CSI估计四大参量，动态地调整传输模式和分配符号资源。为了获得最优的传输策略，将多包合并传输HARQ系统的传输过程建模为马尔可夫决策过程，利用值迭代算法解决了符号分配参数的优化问题。该方法对提高HARQ系统的吞吐率性能具有重要意义。The embodiment of the present invention provides a channel resource allocation method in a multi-packet combined transmission HARQ system. The method introduces real-time CSI estimation feedback. The system estimates four parameters according to confirmation information, transmission rounds, accumulated mutual information and real-time CSI estimation. Dynamically adjust the transmission mode and allocate symbol resources. In order to obtain the optimal transmission strategy, the transmission process of the multi-packet combined transmission HARQ system is modeled as a Markov decision process, and the optimization problem of the symbol allocation parameters is solved by using the value iteration algorithm. This method is of great significance to improve the throughput performance of HARQ system.

与现有的基于累计互信息的技术相比，本发明实现了对CSI估计的利用，基于该指标对时隙的信道资源分配进行了实时的调节。在时隙信道资源固定的情况下，大幅提高了多包合并传输HARQ系统的信道资源利用率。本发明对提高HARQ系统的吞吐率性能，优化时隙信道资源分配具有重要意义，在无线通信系统中具有广阔的应用前景。Compared with the existing technology based on accumulating mutual information, the present invention realizes the utilization of CSI estimation, and adjusts the channel resource allocation of the time slot in real time based on the index. Under the condition that the time slot channel resources are fixed, the channel resource utilization rate of the multi-packet combined transmission HARQ system is greatly improved. The invention has important significance for improving the throughput performance of the HARQ system and optimizing the allocation of time slot channel resources, and has broad application prospects in the wireless communication system.

以下通过一些具体实施例对本发明实施例提供的多包合并传输HARQ系统中的信道资源分配方法的实施方式进行详细说明。The following will describe in detail the implementation manner of the channel resource allocation method in the multi-packet combined transmission HARQ system provided by the embodiments of the present invention through some specific embodiments.

如图1所示，该具体实施例提供的多包合并传输HARQ系统中的信道资源分配方法包括如下具体步骤：As shown in FIG. 1 , the method for channel resource allocation in a multi-packet combined transmission HARQ system provided by this specific embodiment includes the following specific steps:

步骤A：将从发送端到接收端的传输信道建模为独立同分布的块衰落信道，该块衰落信道的信道增益满足瑞利分布；另外，根据香农定理计算块衰落信道的信道容量c的概率分布，将信道容量估计值

与实际值间误差e的概率分布建模为零均值正态分布；其中，每个时隙的估计误差独立同分布；Step A: Model the transmission channel from the sender to the receiver as an independent and identically distributed block fading channel, and the channel gain of the block fading channel satisfies the Rayleigh distribution; in addition, calculate the probability of the channel capacity c of the block fading channel according to Shannon's theorem distribution, the channel capacity estimate

The probability distribution of the error e between the actual value and the actual value is modeled as a zero-mean normal distribution; wherein, the estimated error of each time slot is independent and identically distributed;

步骤B：在第n个时隙开始前，接收端的第二HARQ控制器将第n-1个时隙的数据包译码结果M[n-1]、累计互信息I[n-1]和对第n个时隙的信道容量估计值

反馈给发射端；Step B: Before the start of the nth time slot, the second HARQ controller at the receiving end decodes the data packet decoding result M[n-1] of the n-1th time slot, the accumulated mutual information I[n-1] and Channel capacity estimate for the nth slot

feedback to the transmitter;

步骤C：在第n个时隙，发射端的第一HARQ控制器根据步骤B中的反馈信息以及第n个时隙传输数据包的传输轮次，计算出第n个隙的信道资源分配比例p(0≤p≤1)；其中，位于发送队列队首的队首数据包(HOL数据包)占据p·N_s个符号，位于发送队列中的第二数据包(HOL-next数据包)占据(1-p)·N_s个符号。Step C: In the nth time slot, the first HARQ controller at the transmitting end calculates the channel resource allocation ratio p of the nth slot according to the feedback information in Step B and the transmission round of the transmission data packet in the nth time slot. (0≤p≤1); wherein, the head data packet (HOL data packet) located at the head of the sending queue occupies p·N_s symbols, and the second data packet (HOL-next data packet) located in the sending queue occupies (1-p)·N_s symbols.

该步骤C中，还将累计互信息、信道容量估计值c和决策值p离散化，通过信道容量值c和信道估计误差e的概率分布计算状态转移概率，采用无限阶段平均准则模型的值迭代算法求解最优的决策策略，获得离散状态与离散决策p的映射关系。In this step C, the accumulated mutual information, the channel capacity estimation value c and the decision value p are also discretized, the state transition probability is calculated through the probability distribution of the channel capacity value c and the channel estimation error e, and the value iteration of the infinite stage average criterion model is adopted. The algorithm solves the optimal decision-making strategy and obtains the mapping relationship between the discrete state and the discrete decision p.

步骤D：当0<p<1时，队首数据包(HOL数据包)和第二数据包(HOL-next数据包)以分时的方式共享时隙信道资源在同一时隙合并传输，当p＝0时，HOL数据包被丢弃，HOL-next数据包独享时隙信道资源传输，当p＝1时，HOL数据包独享信道资源完成重传。Step D: When 0<p<1, the first data packet (HOL data packet) and the second data packet (HOL-next data packet) share the time slot channel resources in a time-sharing manner and are combined and transmitted in the same time slot. When p=0, the HOL data packet is discarded, and the HOL-next data packet exclusively uses the time slot channel resource for transmission. When p=1, the HOL data packet exclusively uses the channel resource to complete the retransmission.

按照上述步骤，在系统标称速率为4bpcu(bits per channel use)、平均信噪比从5dB到30dB的情况下，图2为本发明实施例提供的基于CSI估计的对多包合并传输HARQ系统的时隙资源分配方法的实施结果与基于AMI方案的结果对比图，所得图形是平均了T＝1×10⁷总时隙的实验结果。由图2可见，在不同的信噪比条件下，本发明方法实施实例得到的系统吞吐率性能比基于AMI的方法要高。According to the above steps, when the nominal rate of the system is 4bpcu (bits per channel use) and the average signal-to-noise ratio is from 5dB to 30dB, FIG. 2 is a HARQ system for combining multiple packets based on CSI estimation provided by an embodiment of the present invention. The comparison diagram between the implementation result of the time slot resource allocation method and the result based on the AMI scheme, the obtained graph is the experimental result of averaging T=1×10⁷ total time slots. It can be seen from FIG. 2 that, under different signal-to-noise ratio conditions, the system throughput performance obtained by the implementation example of the method of the present invention is higher than that of the method based on AMI.

虽然本发明所揭露的实施方式如上，但所述的内容仅为便于理解本发明而采用的实施方式，并非用以限定本发明。任何本发明所属领域内的技术人员，在不脱离本发明所揭露的精神和范围的前提下，可以在实施的形式及细节上进行任何的修改与变化，但本发明的专利保护范围，仍须以所附的权利要求书所界定的范围为准。Although the embodiments disclosed in the present invention are as above, the described contents are only the embodiments adopted to facilitate the understanding of the present invention, and are not intended to limit the present invention. Any person skilled in the art to which the present invention belongs, without departing from the spirit and scope disclosed by the present invention, can make any modifications and changes in the form and details of the implementation, but the scope of the patent protection of the present invention still needs to be The scope defined by the appended claims shall prevail.

Claims (9)

1. A channel resource allocation method in a multi-packet combining transmission HARQ system is characterized by comprising the following steps:

step A: modeling a transmission channel from a sending end to a receiving end into independent and identically distributed block fading channels, calculating the probability distribution of the channel capacity c of the block fading channels, and estimating the channel capacity

Modeling the probability distribution of the error e between the actual value and the actual value into zero-mean normal distribution; wherein, the estimation error of each time slot is independently and equally distributed;

and B: before the nth time slot starts, the second HARQ controller of the receiving end decodes the data packet of the nth-1 time slot into a result M [ n-1 ]]Accumulating mutual information I [ n-1 ]]And channel capacity estimation for the nth time slot

Feeding back to the transmitting end;

and C: in the nth time slot, the first HARQ controller of the transmitting end calculates the channel resource allocation proportion p (p is more than or equal to 0 and less than or equal to 1) of the nth time slot according to the feedback information in the step B and the transmission round of the data packet transmitted by the nth time slot; discretizing accumulated mutual information, a channel capacity estimation value c and a decision value p, calculating a state transition probability through the probability distribution of the channel capacity value c and a channel estimation error e, and solving an optimal decision strategy by adopting a value iteration algorithm of an infinite-stage average criterion model to obtain a mapping relation between a discrete state and a discrete decision p;

step D: when 0< p <1, the first queue data packet (HOL data packet) and the second data packet (HOL-next data packet) share the time slot channel resource in a time sharing mode and are transmitted in the same time slot in a combined mode, when p is 0, the HOL data packet is discarded, the HOL-next data packet shares the time slot channel resource for transmission, and when p is 1, the HOL data packet shares the time slot channel resource for completing retransmission.

2. The method of claim 1, wherein the channel resource allocation in the multi-packet combining transmission HARQ system,

each time slot transmission data packet comprises the following steps: a head of queue data packet (HOL data packet) located at the head of the transmission queue and a second data packet (HOL-next data packet) located in the transmission queue; wherein the HOL packet occupies p.N_sOne symbol, the HOL-next packet occupying (1-p). N_sA symbol.

3. The method as claimed in claim 2, wherein the step a comprises:

step A1, modeling the transmission channel as independent Rayleigh fading channel with same distribution, then the channel gain h obeys Rayleigh distribution, setting the power of the transmitting signal as epsilon_bPower of channel noise of N₀Then, the signal-to-noise ratio γ at the receiving end is expressed as:

wherein, gamma obeys chi²The probability density function of the distribution, the signal-to-noise ratio γ at the receiving end, is expressed as:

wherein,

representing the average signal-to-noise ratio of the receiving end;

step a2, according to the shannon formula, calculating the maximum mutual information that the receiving end can obtain from each symbol as:

c＝log₂(1+γ)；

wherein c is channel capacity, which represents the maximum rate at which information can be transmitted without errors; obtaining the probability distribution of the channel capacity c through a probability density formula of a composite random variable;

step A3, estimating the channel capacity

Modeling the probability distribution of the error e between the actual value and the normal distribution of the zero mean value and setting

The relationship with c satisfies the following condition:

where e denotes an estimation error of the channel capacity.

4. The method as claimed in claim 3, wherein in step A,

setting the estimation errors e of the channel capacities of different time slots to be mutually uncorrelated, wherein the estimation errors of all time slots are subject to the same zero-mean positive distribution, namely

σ denotes the standard deviation of the estimation error.

5. The method as claimed in claim 4, wherein the acknowledgement information of the data packet decoding result has two values: ACK/NACK; the step B comprises the following steps:

step B1, using u_lAnd u_l+1HOL data packet and HOL-next data packet respectively representing current time slot, l represents data packet number and is used

Representing a data packet u_lAt the k-th_lAcknowledgement information after round transmission; and according to the channel resource division of the (n-1) th time slotThe proportion p feeds back the confirmation information of the data packet decoding result;

step B2, using

Representing a data packet u_lAt the k-th_lThe accumulated mutual information required by the receiving end after the round of transmission is calculated according to a retransmission mechanism;

step B3, the receiving end measures the channel capacity estimation of the nth time slot just before the nth time slot starts

And will be

And feeding back the data packet decoding result and the accumulated mutual information of the (n-1) th time slot to the transmitting end.

6. The method as claimed in claim 5, wherein in step B1,

when 0 is present<p<1, called TS mode, the receiving end has two data packet decoding results, i.e. the acknowledgement information of two data packets is fed back simultaneously

And

when P is 1, the mode is called 1P mode, and only the confirmation information of the data packet is fed back

When P is 0, the mode is called 0P mode, and only the confirmation information of the data packet is fed back

Using M [ n ] to represent the packet decoding result in the acknowledgment message fed back by the nth slot, then M [ n ] is represented as:

7. the method as claimed in claim 6, wherein in step B2,

based on the situation that at most one retransmission is allowed, in TS mode, the HOL packet must be the retransmission packet, and the HOL packet must be discarded in the next time slot, then the cumulative mutual information I [ n ] in the feedback information is represented as:

cumulative mutual information of HOL packets for nth time slot

Accumulated mutual information with HOL-next packets

The accumulated mutual information of the previous time slot is obtained by recursion, and the recursion relation is as follows:

wherein n is_lAnd n_l+1Respectively representing data packets u_lAnd a data packet u_l+1The slot number of the initial transmission.

8. The method as claimed in claim 7, wherein the step C comprises:

step C1, mapping the feedback information and the value of the transmission turn in the step B to a corresponding discrete interval to obtain the discrete state of the current HARQ system;

and step C2, searching the p value of the optimal decision strategy according to the mapping relation from the discrete state to the discrete decision p calculated in the step C1.

9. The method as claimed in claim 8, wherein the step D comprises:

step D1, determining the current transmission mode according to the found p value of the optimal decision strategy;

step D2, 1P mode, entire transmission time slot N_sAll symbols are allocated to subcodes of the HOL packet;

in the 0P mode, all symbol resources of the whole transmission time slot are allocated to sub-code words of the HOL-next data packet;

TS mode, symbols of whole transmission time slot are shared by sub-code words of two data packets of HOL and HOL-next, and occupy p.N_sA symbol sum (1-p). N_sAnd the information bits of the two data packets do not interfere with each other during coding, and the symbol mapping during modulation does not overlap with each other.

Images