CN103796317A

Movatterモバイル変換

Info

Publication number: CN103796317A
Application number: CN201310654633.XA
Authority: CN
Inventors: 孙君; 陶宇; 邵世祥
Original assignee: Nanjing Post and Telecommunication University
Current assignee: Jiangsu Zhongcheng Communication Technology Co Ltd
Priority date: 2013-12-06
Filing date: 2013-12-06
Publication date: 2014-05-14
Anticipated expiration: 2033-12-06
Also published as: CN103796317B

Abstract

The invention relates to a device-to-device relay communication-based resource allocation method. In a single cell, an LTE-Advanced cellular network and a device-to-device (D2D) system constitute a hybrid network, and the duplex mode of the cellular network is a time division duplex mode, and the device-to-device (D2D) system utilizes the uplink resources of the cellular network in the cell in a multiplex mode. The device-to-device relay communication-based resource allocation method includes the steps of interaction with a base station, update of a resource record table, application and distribution of resources, and selection of relay nodes. According to the device-to-device relay communication-based resource allocation method of the invention, at first, system resources are allocated to active cellular subscribers according to polling standards under the premise that communication quality can be ensured, and then, most suitable resources are selected for device-to-device (D2D) subscribers in the cell through using a genetic algorithm according to maximum resource utilization rate standards. The objective of the invention is to minimize the probability of resource allocation failure and enable more device-to-device (D2D) subscribers can perform communication with normal communication of the cellular subscribers ensured.

Description

Translated fromChinese

一种基于终端直通中继通信的资源分配方法A resource allocation method based on terminal direct relay communication

技术领域technical field

本发明涉及一种基于终端直通(D2D)中继通信的资源分配方法，利用LTE-Advanced蜂窝网络、D2D通信和中继技术，通过基站进行集中式的资源调度，适用于未来的混合系统(Hybrid System)。 The invention relates to a resource allocation method based on direct-to-device (D2D) relay communication, which uses LTE-Advanced cellular network, D2D communication and relay technology to perform centralized resource scheduling through base stations, and is suitable for future hybrid systems (Hybrid System). the

背景技术Background technique

近年来，第三代合作伙伴计划(3GPP)长期演进(LTE)技术因其展示出的高数据速率、高系统容量等优势，得到了全球范围内的瞩目，大量的研究力量已经或正在为它的发展做贡献。可以预见，LTE技术将是实现IMT-Advanced技术提案的坚实基础，并且LTE技术研究现已步入了LTE-Advanced阶段。 In recent years, the 3rd Generation Partnership Project (3GPP) Long Term Evolution (LTE) technology has attracted worldwide attention because of its advantages such as high data rate and high system capacity, and a large number of research forces have been or are working on it contribute to the development of It can be predicted that LTE technology will be a solid foundation for realizing the IMT-Advanced technical proposal, and LTE technology research has entered the LTE-Advanced stage. the

对于蜂窝机制来说，主要优点在于便捷的资源分配和功率控制，同时明显的缺点在于对无线频谱资源的未充分利用，而无线频谱紧张的问题又日益严峻。针对这一情况，一种新的无线接口技术，终端直通(Device-to-Device,D2D)技术被提出，以形成一种新的通信模式，一经提出便得到广泛关注。D2D不仅工作在授权频段，并且接受基站的协助调度，具有低成本、灵活性，可以带来高效率的频谱利用，缓解基站负载，降低能源消耗。未来的混合系统(Hybrid System)中，D2D通信将是一个重要的组成部分。 For the cellular mechanism, the main advantage lies in convenient resource allocation and power control, while the obvious disadvantage lies in the underutilization of wireless spectrum resources, and the problem of wireless spectrum shortage is becoming more and more severe. In view of this situation, a new wireless interface technology, Device-to-Device (D2D) technology is proposed to form a new communication mode, and it has attracted widespread attention once it is proposed. D2D not only works in the licensed frequency band, but also accepts the assisted scheduling of the base station. It has low cost and flexibility, can bring high-efficiency spectrum utilization, alleviate the load of the base station, and reduce energy consumption. In the future hybrid system (Hybrid System), D2D communication will be an important component. the

如果D2D发送端和接收端之间相距过远，或者他们之间的直接链路信道质量不够好，不足以进行D2D通信，这时候就需要借助中继(Relay)来实现D2D通信。但是，利用D2D用户附近的空闲用户来充当移动中继进行D2D通信，将会有多种干扰链路混合，使得干扰问题变得更加复杂，同时，由于中继系统需要采用2个传输时间间隔(TTI)来进行一次传输，因而相较于直连模式，吞吐量会不可避免的有所下降，因此资源分配的目标应集中在降低D2D用户资源分配失败概率，提高系统的总体应用性和适用性。 If the distance between the D2D sending end and the receiving end is too far, or the quality of the direct link channel between them is not good enough for D2D communication, then a relay (Relay) is needed to realize D2D communication. However, if idle users near D2D users are used as mobile relays for D2D communication, there will be a mixture of multiple interference links, making the interference problem more complicated. At the same time, since the relay system needs to use two transmission time intervals ( TTI) to perform a transmission, so compared with the direct connection mode, the throughput will inevitably decrease, so the goal of resource allocation should focus on reducing the probability of D2D user resource allocation failure and improving the overall applicability and applicability of the system. . the

发明内容Contents of the invention

本发明的目的是提供一种基于终端直通(D2D)中继通信的资源分配方法，在D2D中继通信的干扰分析基础上，提供一种可靠的高性能的基于LTE-Advanced蜂窝网络的D2D中继通信的资源分配方案，以降低D2D资源分配失败的概率，使更多的D2D用户可以进行通信，提高系统的总体适用性。 The purpose of the present invention is to provide a resource allocation method based on direct-to-device (D2D) relay communication, based on the interference analysis of D2D relay communication, to provide a reliable high-performance D2D medium based on LTE-Advanced cellular network A resource allocation scheme for communication is adopted to reduce the probability of D2D resource allocation failure, so that more D2D users can communicate and improve the overall applicability of the system. the

本发明的技术方案如下：一种基于终端直通中继通信的资源分配方法，其特征在于：在单个小区中LTE-Advanced蜂窝网络与D2D系统构成混合网络，蜂窝网络的双工方式为时分双工，D2D以复用的模式利用小区内蜂窝通信的上行资源，包括与基站的交互、资源记录表的更新、资源的申请与分配、中继节点的选取过程；首先根据轮询准则，保证通信质量为前提，将系统资源分配给活跃蜂窝用户，然后根据最大化资源利用率准则，利用遗传算法来为小区中的D2D用户选择最合适的资源进行复用，目标是最小化资源分配失败的概率，在保证蜂窝用户正常通信的基础上，使更多的D2D用户可以进行通信；包括以下步骤： The technical solution of the present invention is as follows: a resource allocation method based on terminal direct relay communication, characterized in that: in a single cell, the LTE-Advanced cellular network and the D2D system form a hybrid network, and the duplex mode of the cellular network is time division duplex , D2D uses the uplink resources of cellular communication in the cell in a multiplexing mode, including the interaction with the base station, the update of the resource record table, the application and allocation of resources, and the selection process of the relay node; firstly, according to the polling criterion, the communication quality is guaranteed As a premise, allocate system resources to active cellular users, and then use genetic algorithm to select the most suitable resources for D2D users in the cell for multiplexing according to the criterion of maximizing resource utilization. The goal is to minimize the probability of resource allocation failure. On the basis of ensuring normal communication of cellular users, enable more D2D users to communicate; including the following steps:

步骤1:假设所有的蜂窝用户以固定功率P_CO进行通信，而所有的D2D用户的发送功率则固定为P_SO，所有的用户终端UE，包括活跃的、空闲的蜂窝用户和D2D用户都一直与基站保持着连接并周期地与基站交换信息，即使是正在进行D2D通信的用户，仍然要受基站控制； Step 1: Assuming that all cellular users communicate with a fixed power P_CO , while the transmission power of all D2D users is fixed at P_SO , all user terminals UE, including active and idle cellular users and D2D users, are always connected with The base station maintains the connection and periodically exchanges information with the base station, even users who are communicating in D2D are still subject to the control of the base station;

步骤2:上行蜂窝信道集合记为

，所有的无线资源分配和调度都由基站完成，基站维持着一个本小区资源分配实时状况的信道记录表，记作ChnTable，用于记录各信道的使用情况，只要资源分配信息发生了变化，基站就会对其进行更新，一个信道内最多同时共存一个蜂窝用户和一对基于移动中继的D2D用户； Step 2: The set of uplink cellular channels is denoted as

, all wireless resource allocation and scheduling are completed by the base station. The base station maintains a channel record table of the real-time resource allocation status of the cell, denoted as ChnTable, which is used to record the usage of each channel. As long as the resource allocation information changes, the base station It will be updated, and at most one cellular user and a pair of mobile relay-based D2D users can coexist in one channel at the same time;

步骤3:N个蜂窝用户C_UEs使用N个不同的上行蜂窝信道与基站进行通信，彼此之间没有干扰，资源分配工作由基站完成，并在ChnTable中记录更新； Step 3: N cellular users C_UEs use N different uplink cellular channels to communicate with the base station, there is no interference between each other, the resource allocation work is completed by the base station, and records are updated in ChnTable;

步骤4:当有潜在的D2D通信可能时，D2D发送用户S_UE和接收用户D_UE分别将它们的位置信息报告给基站，这一过程利用内置的全球定位系统GPS或者北斗卫星导航系统BDS模块来完成，并根据干扰模型来进行蜂窝用户和D2D用户的信号干扰噪声比SINR计算，由基站来决定是否准许进行D2D尝试，如果不合适的话，该对D2D用户会被基站拒绝并进入等待队列； Step 4: When potential D2D communication is possible, D2D sending user S_UE and receiving user D_UE report their location information to the base station respectively. This process is completed using the built-in global positioning system GPS or Beidou satellite navigation system BDS module, And according to the interference model to calculate the SINR of the cellular user and the D2D user, the base station decides whether to allow the D2D attempt. If it is not suitable, the pair of D2D users will be rejected by the base station and enter the waiting queue;

首先由基站检查ChnTable，去除已经饱和即一个蜂窝用户和一对D2D用户共存的信道，剩余信道集合记为

，如果其不为空的话，分别对其中的各个信道进行进一步分析，看能否容纳下该对D2D用户使得该信道下所有用户都能满足通信要求，筛选出的合格信道即蜂窝用户SINR达到蜂窝用户的目标SINRγ_C-target，D2D用户SINR达到D2D用户的目标SINRγ_D-target，组成集合，并根据改进的遗传算法从

中选出目标信道供该D2D对复用，遗传算法的进化目标选取的是最小化资源分配失败的概率； Firstly, the base station checks the ChnTable, and removes channels that have been saturated, that is, a cellular user and a pair of D2D users coexist, and the remaining channel set is denoted as

, if it is not empty, further analyze each of the channels to see if the pair of D2D users can be accommodated so that all users under the channel can meet the communication requirements. The target SINRγ_C-target of the user, and the SINR of the D2D user reaches the target SINRγ_D-target of the D2D user, forming a set , and according to the improved genetic algorithm from

Select the target channel for the D2D pair to reuse, and the evolutionary goal of the genetic algorithm is to minimize the probability of resource allocation failure;

如果没有找到可行信道，则该队D2D用户将被拒绝并进入等待队列； If no feasible channel is found, the team of D2D users will be rejected and enter the waiting queue;

步骤5:进行移动中继的选择，以D2D用户对收发端连线为直径构成的圆形区域中总有若干空闲用户可供选取作为移动中继，从中为每对D2D用户选取m个空闲用户作为候选中继集合，一个空闲用户只会出现在一对D2D用户的候选中继集合中，同一时刻一个空闲用户只能为一对D2D用户服务，一对D2D用户只能选择一个空闲用户作为移动中继；依据D2D中继系统SINR最优原则，并根据干扰模型来进行蜂窝用户和D2D用户的SINR计算，从候选移动中继集合中选取出移动中继，使得该D2D中继通信可以获得最大的SINR；若某个空闲用户电量不足或不适合用于移动中继，则从候选中继集合中去除该空闲用户； Step 5: Carry out the selection of mobile relays, there are always some idle users available for selection as mobile relays in the circular area formed by the diameter of the D2D user's connection to the transmitting and receiving end, from which m idle users are selected for each pair of D2D users As a candidate relay set, an idle user will only appear in the candidate relay set of a pair of D2D users. At the same time, an idle user can only serve a pair of D2D users, and a pair of D2D users can only choose one idle user as the mobile Relay: According to the principle of optimal SINR of the D2D relay system, and according to the interference model to calculate the SINR of the cellular user and the D2D user, select the mobile relay from the candidate mobile relay set, so that the D2D relay communication can obtain the maximum SINR; if an idle user has insufficient power or is not suitable for mobile relay, remove the idle user from the candidate relay set;

步骤6:利用步骤4选出的信道和步骤5选出的移动中继，开始进行D2D测试通信，如果实际测试表明该信道内蜂窝通信和D2D中继通信都能够得到保障，则允许进行正式的D2D中继通信并且基站更新ChnTable的信息记录；否则，返回步骤4重新进行信道选择，直到通过测试； Step 6: Use the channel selected in step 4 and the mobile relay selected in step 5 to start D2D test communication. If the actual test shows that both the cellular communication and the D2D relay communication in the channel can be guaranteed, the formal communication is allowed. D2D relay communication and the base station updates the information record of ChnTable; otherwise, return to step 4 and re-select the channel until the test is passed;

步骤7:一旦有用户完成了自己的通信，或者出现可利用的信道资源，ChnTable中的信息将会被立即更新并依次序重新对等待队列中的D2D用户执行步骤4；为了保证持续可靠的通信过程，基站会周期对ChnTable中记录的用户所进行的通信进行监视。 Step 7: Once a user completes his own communication, or available channel resources appear, the information in ChnTable will be updated immediately and step 4 will be re-executed for the D2D users in the waiting queue in order; in order to ensure continuous and reliable communication During the process, the base station periodically monitors the communications performed by the users recorded in the ChnTable. the

所述的遗传算法包括如下步骤： Described genetic algorithm comprises the steps:

步骤一:对基因进行实数编码；染色体长度为蜂窝信道数N，共含有N个基因，对基因进行实数编码，一个信道最多只能被一对D2D用户复用，一对D2D用户最多只能复用一个信道，某个基因值为0表示该信道没有被D2D用户复用，第i个基因的数值A_i表示第A_i对D2D用户复用的是第i个信道，因而基因型A＝A₁A₂…A_N所对应的表现型为x＝(x₁,x₂,…,x_N)^H；即： Step 1: Encode the genes with real numbers; the length of the chromosome is the number of cellular channels N, which contains N genes in total. The genes are encoded with real numbers. A channel can only be multiplexed by a pair of D2D users at most, and a pair of D2D users can only multiplex at most. Using a channel, a gene value of 0 indicates that the channel is not reused by D2D users, and the value A_i of the i-th gene indicates that the i_- th channel is multiplexed by the A i-th pair of D2D users, so genotype A=A₁ A₂ …A_N corresponds to x=(x₁ ,x₂ ,…,x_N )^H ; that is:

A_i＝x_iA_i = x_i

显然，A_i＝0表示该信道没有被任何一对D2D用户复用，而对于

，如果A_i≠0且A_j≠0，有A_i≠A_j； Obviously, A_i =0 means that the channel is not multiplexed by any pair of D2D users, and for

, if A_i ≠0 and A_j ≠0, then A_i ≠A_j ;

步骤二:对当前代数的种群进行选择操作；由于进化目标选取的是最小化资源分配失败的概率，因此用D2D用户的资源分配成功率来评价个体的适应度，如下式，用于选择操作，其中

表示第l代群体中第i个个体； Step 2: Select the population of the current generation; since the evolution goal is to minimize the probability of resource allocation failure, the resource allocation success rate of the D2D user is used to evaluate the fitness of the individual, as follows, for the selection operation, in

Indicates the i-th individual in the l-generation population;

依据基于上限的确定式采样，降低因选择的偶然性而造成退化的可能，确保选择出最优异的个体；群体个数用Q来表示： According to the deterministic sampling based on the upper limit, the possibility of degradation caused by the chance of selection is reduced, and the best individuals are selected; the number of groups is represented by Q:

(3)计算个体在下一代群体中的期望生存数目，用Q_i来表示，如下式： (3) Calculate the expected survival number of individuals in the next generation group, expressed by_Qi , as follows:

${Q Q}_{i i} = = Q Q \cdot &Center Dot; F f (({A A}_{i i}^{l l})) / / {Σ Σ}_{i i = = 11}^{Q Q} F f (({A A}_{i i}^{l l}))$

(4)确定各对应个体在下一代群体中的准生存数目，

向上取整，从而得到准下一代群体的个体

个； (4) Determine the quasi-survival number of each corresponding individual in the next generation group,

Round up to get the individuals of the quasi-next generation group

indivual;

(3)从准下一代群体中删去适应度最低的

个个体，从而得到Q个经过选择的个体； (3) Delete the one with the lowest fitness from the quasi-next generation group

individuals, so as to obtain Q selected individuals;

步骤三:对经过选择操作的种群进行行内交叉；采用行内交叉的操作，对经过选择操作的种群进行行内交叉，确保生成的新个体仍是有效的信道排列，能够作为一种信道分配方案； Step 3: Carry out intra-row crossover to the population through selection operation; adopt the operation of intra-row crossover, carry out intra-row crossover to the population through selection operation, ensure that the new individual generated is still an effective channel arrangement, and can be used as a channel allocation scheme;

先依从交叉概率p_c对经过选择操作后的群体随机选取出要交叉的个体，但是不再对其进行两两配对，而是对每个选出来的待交叉个体，都由1到N之间随机生成两个随机数i、j(i≠j)，并交换A_i和A_j的数值，如果新产生的个体满足可行解要求，新个体，则予以保留，若不满足取值要求，则删除该个体，并继续通过交叉操作产生新个体直到满足可行解要求为止； First, according to the crossover probability p_c, randomly select individuals to be crossover from the group after the selection operation, but no pairwise pairing is performed on them, but for each selected individual to be crossover, each selected crossover individual is between 1 and N Randomly generate two random numbers i and j (i≠j), and exchange the values of A_i and A_j . If the newly generated individual meets the requirements of feasible solutions, the new individual will be retained. If the value requirement is not met, then Delete the individual, and continue to generate new individuals through the crossover operation until the requirements of the feasible solution are met;

步骤四:对经过交叉操作的种群进行均匀变异，确保能够生成新的信道分配方案； Step 4: uniformly mutating the cross-operated population to ensure that a new channel allocation scheme can be generated;

先依从变异概率p_m从经过交叉后的群体中随机选出需要进行变异的个体，对每一个需要进行变异的个体，由1到N之间随机生成一个随机数k，对基因A_k进行变异操作，从{1,...,N}-{A_i|i＝1,...,N}这个差集中随机选出一个元素，将其赋给A_k作为突变，若这个差集为空，则将其置为0。若新产生个体满足解空间的参数取值要求，新个体将予以保留，若不满足解空间参数取值要求，则删除该个体，并继续通过变异产生新个体直到满足参数取值要求为止； Firstly, according to the mutation probability p_m , randomly select individuals who need to be mutated from the population after crossover, and for each individual who needs to be mutated, a random number k is randomly generated from 1 to N, and the gene A_k is mutated Operation, randomly select an element from the difference set {1,...,N}-{A_i |i=1,...,N}, and assign it to A_k as a mutation, if the difference set is If empty, set it to 0. If the newly generated individual meets the parameter value requirements of the solution space, the new individual will be retained; if it does not meet the parameter value requirements of the solution space, the individual will be deleted, and new individuals will continue to be generated through mutation until the parameter value requirements are met;

步骤五:重复步骤2-步骤4，直到进化到最终代数T，从中选取出适应度最高的个体作为资源分配方案。 Step 5: Repeat step 2-step 4 until the final algebra T is evolved, and select the individual with the highest fitness as the resource allocation scheme. the

所述步骤4和步骤5中根据干扰模型来进行蜂窝用户和D2D用户的SINR计算，包括如下步骤： In the step 4 and step 5, the SINR calculation of the cellular user and the D2D user is performed according to the interference model, including the following steps:

步骤A:D2D中继通信模型中，蜂窝用户C正在与基站B进行上行通信，S-R-D是一个D2D中继通信系统，复用的是C占用的无线资源，S表示D2D发送用户，R表示移动中继，D表示D2D接收用户。有R、D、B三处接收到的混合信号分别为： Step A: In the D2D relay communication model, cellular user C is conducting uplink communication with base station B. S-R-D is a D2D relay communication system that multiplexes the wireless resources occupied by C. S represents the D2D sending user, and R represents the mobile Next, D represents a D2D receiving user. The mixed signals received by R, D, and B are respectively:

${y the y}_{R R} = = {h h}_{SR SR} \sqrt{{P P}_{S S} {d d}_{SR SR}^{- - α α}} {x x}_{S S} + + {h h}_{CR CR} \sqrt{{P P}_{C C} {d d}_{CR CR}^{- - α α}} {x x}_{C C} + + {n no}_{R R}$

${y the y}_{D D.} = = {h h}_{RD RD} \sqrt{{P P}_{R R} {d d}_{RD RD}^{- - α α}} {x x}_{R R} + + {h h}_{CD cd} \sqrt{{P P}_{C C} {d d}_{CD cd}^{- - α α}} {x x}_{C C} + + {n no}_{D D.}$

x_S、x_R、x_C分别为对应的发送信号，P_S、P_R、P_C为对应的发送功率，n_R、n_D、n_B代表对应的加性高斯白噪声AWGN，相应的噪声功率分别为N_R、N_D、N_B，α表示路径损耗指数；另外，h_XY、d_XY分别用于表示X-Y链路的信道系数、通信距离。 x_S , x_R , x_C are the corresponding transmission signals,_PS ,_PR ,_PC are the corresponding transmission powers, n_R , n_D , n_B are the corresponding additive white Gaussian noise AWGN, and the corresponding noise The powers are_NR ,_ND , and_NB , and α represents the path loss index; in addition, h_XY and d_XY are used to represent the channel coefficient and communication distance of the XY link, respectively.

步骤B:两种中继模式，AF和DF模式，两者的区别在这里主要体现在x_R表示式的不同，根据不同的中继模式来计算D2D用户的SINR： Step B: Two relay modes, AF mode and DF mode, the difference between the two is mainly reflected in the expression of x_R , and the SINR of the D2D user is calculated according to different relay modes:

AF模式下，R仅仅是将接收到的信号进行放大转发，并不进行信息解码。 In AF mode, R only amplifies and forwards the received signal without decoding information. the

$\begin{matrix} {x x}_{R R}^{AF AF} = = {ky ky}_{R R} = = \sqrt{\frac{{P P}_{R R}}{{| | {h h}_{SR SR} | |}^{22} {P P}_{S S} {d d}_{SR SR}^{- - α α} + + {| | {h h}_{CR CR} | |}^{22} {P P}_{C C} {d d}_{CR CR}^{- - α α} + + {N N}_{R R}}} {y the y}_{R R} \\ = = \sqrt{\frac{{P P}_{R R}}{{| | {h h}_{SR SR} | |}^{22} {P P}_{S S} {d d}_{SR SR}^{- - α α} + + {| | {h h}_{CR CR} | |}^{22} {P P}_{C C} {d d}_{CR CR}^{- - α α} + + {N N}_{R R}}} (({h h}_{SR SR} \sqrt{{P P}_{S S} {d d}_{SR SR}^{- - α α}} {x x}_{S S} + + {h h}_{CR CR} \sqrt{{P P}_{C C} {d d}_{CR CR}^{- - α α}} {x x}_{C C} + + {n no}_{R R})) \end{matrix}$

$k k = = \sqrt{\frac{{P P}_{R R}}{{| | {h h}_{SR SR} | |}^{22} {P P}_{S S} {d d}_{SR SR}^{- - α α} + + {| | {h h}_{CR CR} | |}^{22} {P P}_{C C} {d d}_{CR CR}^{- - α α} + + {N N}_{R R}}} = = \sqrt{\frac{{P P}_{R R}}{{G G}_{SR SR} {P P}_{S S} + + {G G}_{CR CR} {P P}_{C C} + + {N N}_{R R}}}$

其中，k代表放大增益，G_XY用来表示X-Y链路的信道增益，即

对x_R进行归一化得到： Among them, k represents the amplification gain, and G_XY is used to represent the channel gain of the XY link, namely

Normalizing x_R gives:

${x x}_{R R}^{AF AF} = = \sqrt{\frac{11}{{| | {h h}_{SR SR} | |}^{22} {P P}_{S S} {d d}_{SR SR}^{- - α α} + + {| | {h h}_{CR CR} | |}^{22} {P P}_{C C} {d d}_{CR CR}^{- - α α} + + {N N}_{R R}}} (({h h}_{SR SR} \sqrt{{P P}_{S S} {d d}_{SR SR}^{- - α α}} {x x}_{S S} + + {h h}_{CR CR} \sqrt{{P P}_{C C} {d d}_{CR CR}^{- - α α}} {x x}_{C C} + + {n no}_{R R}))$

最终得到R处的接收信号、SINR表达式分别为： Finally, the received signal at R and the SINR expressions are:

$\begin{matrix} {y the y}_{D D.}^{AF AF} = = {kh kh}_{SR SR} \sqrt{{P P}_{S S} {d d}_{SR SR}^{- - α α}} {h h}_{RD RD} \sqrt{{d d}_{RD RD}^{- - α α}} {x x}_{S S} + + k k {h h}_{CR CR} \sqrt{{P P}_{C C} {d d}_{CR CR}^{- - α α}} {h h}_{RD RD} \sqrt{{d d}_{RD RD}^{- - α α}} {x x}_{C C} \\ + + k k {h h}_{RD RD} \sqrt{{d d}_{RD RD}^{- - α α}} {n no}_{R R} + + {h h}_{CD cd} \sqrt{{P P}_{C C} {d d}_{CD cd}^{- - α α}} {x x}_{C C} + + {n no}_{D D.} \end{matrix}$

$\begin{matrix} {γ γ}_{D D.}^{AF AF} = = \frac{{k k}^{22} {| | {h h}_{SR SR} | |}^{22} {d d}_{SR SR}^{- - α α} {| | {h h}_{RD RD} | |}^{22} {d d}_{RD RD}^{- - α α} {P P}_{S S}}{} \\ = = \frac{{k k}^{22} {G G}_{SR SR} {G G}_{RD RD} {P P}_{S S}}{{k k}^{22} {G G}_{CR CR} {G G}_{RD RD} {P P}_{C C} + + {k k}^{22} {G G}_{RD RD} {N N}_{R R} + + {G G}_{CD cd} {P P}_{C C} + + {N N}_{D D.}} \end{matrix}$

DF模式下，R处涉及到一个能否正确解码的问题： In DF mode, R is related to a problem of whether it can be decoded correctly:

这里假设R处总可以正确解码出x_S，从而D处接收信号，以及R、D处各自的SINR为： Here it is assumed that x_S can always be correctly decoded at R, so that the received signal at D, and the respective SINRs at R and D are:

${y the y}_{D D.}^{DF DF} = = {h h}_{RD RD} \sqrt{{P P}_{R R} {d d}_{RD RD}^{- - α α}} {x x}_{S S} + + {h h}_{CD cd} \sqrt{{P P}_{C C} {d d}_{CD cd}^{- - α α}} {x x}_{C C} + + {n no}_{D D.}$

${γ γ}_{R R}^{DF DF} = = \frac{{G G}_{SR SR} {P P}_{S S}}{{G G}_{CR CR} {P P}_{C C} + + {N N}_{R R}}$

${γ γ}_{D D.}^{DF DF} = = \frac{{G G}_{RD RD} {P P}_{R R}}{{G G}_{CD cd} {P P}_{C C} + + {N N}_{D D.}}$

步骤C:最大比合并MSC；对于D2D接收端的最终信号接收，采用基于最大比的分集合并方式，在D处进行两个支路信号的最大比合并，并且分析可知最大比合并后的总的SINR为各支路SINR之和： Step C: The maximum ratio combining MSC; for the final signal reception of the D2D receiving end, the maximum ratio combining method based on the maximum ratio is adopted, and the maximum ratio combining of the two branch signals is performed at D, and the analysis shows that the total SINR after the maximum ratio combining is the sum of the SINR of each branch:

${γ γ}_{D D.} = = {γ γ}_{D D.}^{direct directly} + + {γ γ}_{D D.}^{relay relay}$

其中，

表示直接链路的SINR，

表示中继链路的SINR： in,

Indicates the SINR of the direct link,

Indicates the SINR of the trunk link:

${γ γ}_{D D.}^{direct directly} = = \frac{{| | {h h}_{SD SD} | |}^{22} {d d}_{SD SD}^{- - α α} {P P}_{S S}}{{| | {h h}_{CD cd} | |}^{22} {d d}_{CD cd}^{- - α α} {P P}_{C C} + + {N N}_{D D.}} = = \frac{{G G}_{SD SD} {P P}_{S S}}{{G G}_{CD cd} {P P}_{C C} + + {N N}_{D D.}}$

步骤D:计算蜂窝用户的SINR；上面的分析都是针对D2D系统，而对于蜂窝用户来说，蜂窝用户的SINR表达式为： Step D: Calculate the SINR of the cellular user; the above analysis is all for the D2D system, and for the cellular user, the SINR expression of the cellular user is:

当S向R发送信号时；

When S sends a signal to R;

当R向D发送信号时。

When R sends a signal to D.

本发明的优点及有益效果：本发明在D2D中继通信的干扰分析基础上，提出的一种基于遗传算法的资源分配方案，降低了D2D资源分配失败的概率，使更多的D2D用户可以进行通信，提升了系统资源的利用率。 Advantages and beneficial effects of the present invention: Based on the interference analysis of D2D relay communication, the present invention proposes a resource allocation scheme based on genetic algorithm, which reduces the probability of D2D resource allocation failure and enables more D2D users to perform Communication improves the utilization of system resources. the

附图说明Description of drawings

图1是D2D中继通信的场景示意图； Figure 1 is a schematic diagram of a scenario of D2D relay communication;

图2是D2D中继通信的干扰分析基本模型； Figure 2 is the basic model of interference analysis of D2D relay communication;

图3是基于最大比合并的D2D中继通信模型； Figure 3 is a D2D relay communication model based on maximum ratio combining;

图4是资源分配的流程示意图； Fig. 4 is a schematic flow chart of resource allocation;

图5本发明中的遗传算法流程示意图； The schematic flow chart of genetic algorithm among Fig. 5 the present invention;

图6是信道资源的分配记录表。 Fig. 6 is an allocation record table of channel resources. the

具体实施方式Detailed ways

单个小区的D2D中继通信系统模型如图1所示，LTE-Advanced蜂窝网络与D2D系统构成混合网络，其中蜂窝网络的双工方式为时分双工(TDD)，D2D以复用的模式利用小区内蜂窝通信的上行资源。主要的干扰是两方面：一方面蜂窝用户发送会对复用同一资源的D2D用户接收造成干扰，对移动中继的接收和D2D接收端的接收都会造成干扰；另一方面D2D发送端的发送和移动中继的发送都会对蜂窝用户造成干扰，也就是干扰基站(BS)对蜂窝信号的接收。 The D2D relay communication system model of a single cell is shown in Figure 1. The LTE-Advanced cellular network and the D2D system constitute a hybrid network. The duplex mode of the cellular network is time division duplex (TDD), and D2D utilizes cell Uplink resources for intracellular communication. The main interference is two aspects: on the one hand, the transmission of cellular users will interfere with the reception of D2D users who multiplex the same resource, and will cause interference to the reception of mobile relays and the reception of D2D receivers; on the other hand, the transmission of D2D transmitters and the mobile The subsequent transmission will cause interference to cellular users, that is, interfere with the reception of cellular signals by the base station (BS). the

考虑单个小区场景，并假设小区间干扰已妥善处理，小区半径为R，基站位于小区中心，N个蜂窝用户(用C_UE₁,C_UE₂,…,C_UE_N来表示)利用N个不同的信道，与基站进行上行通信，彼此之间没有干扰。假设每个蜂窝用户占用的信道最多被1对D2D用户复用，且D2D通信和蜂窝通信是同步的，N对D2D用户(D2D发送端:S_UE₁,S_UE₂,…,S_UE_N;D2D接收端:D_UE₁,D_UE₂,…,D_UE_N)复用所有的资源块。 Consider a single cell scenario, and assume that the inter-cell interference has been properly handled, the cell radius is R, the base station is located in the center of the cell, and N cellular users (denoted by C_UE₁ , C_UE₂ ,...,C_UE_N ) use N different channels, Uplink communication with the base station without interference between each other. Assuming that the channel occupied by each cellular user is multiplexed by at most one pair of D2D users, and D2D communication and cellular communication are synchronized, N pairs of D2D users (D2D sending end: S_UE₁ , S_UE₂ ,..., S_UE_N ; D2D receiving end : D_UE₁ , D_UE₂ ,..., D_UE_N ) multiplex all resource blocks.

由于上行链路采用的无线接入技术是SC-FDMA，每个蜂窝用户进行通信至少需要2个资源块(resource blocks,RBs)，并且必须在频域上连续，假设N个蜂窝信道均由两个频域上连续的资源块组成。 Since the wireless access technology used in the uplink is SC-FDMA, each cellular user needs at least 2 resource blocks (resource blocks, RBs) for communication, and must be continuous in the frequency domain, assuming that N cellular channels are composed of two Consecutive resource blocks in the frequency domain. the

除此之外，还有M个空闲的蜂窝用户(用I_UE₁,I_UE₂,…,I_UE_M来表示)，这M个空闲用户可以充当移动中继。 In addition, there are M idle cellular users (indicated by I_UE₁ , I_UE₂ , . . . , I_UE_M ), and these M idle users can act as mobile relays.

如图1所示，C_UE₁与基站之间正在进行上行蜂窝通信，S_UE₁与D_UE₁借助I_UE₂作为移动中继进行D2D中继通信，复用的是C_UE₁占用的资源，值得注意的是，需要在距离基站较近的区域内禁止进行D2D通信，以免对蜂窝用户产生不可协调的干扰。 As shown in Figure 1, the uplink cellular communication between C_UE₁ and the base station is in progress. S_UE₁ and D_UE₁ use I_UE₂ as a mobile relay to carry out D2D relay communication, and the resources occupied by C_UE₁ are multiplexed. It is worth noting that , it is necessary to prohibit D2D communication in an area close to the base station, so as to avoid uncoordinated interference to cellular users.

对D2D中继通信进行干扰建模，干扰分析基本模型如图2所示，蜂窝用户C正在与基站B进行上行通信，S-R-D是一个D2D中继通信系统，复用的是C占用的无线资源，S表示D2D发送用户，R表示移动中继，D表示D2D接收用户。有R、D、B三处接收到的混合信号分别为： Interference modeling is performed on D2D relay communication. The basic model of interference analysis is shown in Figure 2. Cellular user C is communicating uplink with base station B. S-R-D is a D2D relay communication system that multiplexes the wireless resources occupied by C. S indicates a D2D sending user, R indicates a mobile relay, and D indicates a D2D receiving user. The mixed signals received by R, D, and B are respectively:

${y the y}_{R R} = = {h h}_{SR SR} \sqrt{{P P}_{S S} {d d}_{SR SR}^{- - α α}} {x x}_{S S} + + {h h}_{CR CR} \sqrt{{P P}_{C C} {d d}_{CR CR}^{- - α α}} {x x}_{C C} + + {n no}_{R R} - - - - - - ((11))$

${y the y}_{D D.} = = {h h}_{RD RD} \sqrt{{P P}_{R R} {d d}_{RD RD}^{- - α α}} {x x}_{R R} + + {h h}_{CD cd} \sqrt{{P P}_{C C} {d d}_{CD cd}^{- - α α}} {x x}_{C C} + + {n no}_{D D.} - - - - - - ((22))$

x_S、x_R、x_C分别为对应的发送信号，P_S、P_R、P_C为对应的发送功率，n_R、n_D、n_B代表对应的加性高斯白噪声(AWGN)，相应的噪声功率分别为N_R、N_D、N_B，α表示路径损耗指数；另外，在本文中，h_XY、d_XY分别用来表示X-Y链路的信道系数、通信距离。 x_S , x_R , x_C are the corresponding transmission signals,_PS ,_PR ,_PC are the corresponding transmission powers, n_R , n_D , n_B represent the corresponding additive white Gaussian noise (AWGN), and the corresponding The noise power of is_NR ,_ND , and_NB , and α represents the path loss index; in addition, in this paper, h_XY and d_XY are used to represent the channel coefficient and communication distance of the XY link, respectively.

两种中继模式，AF和DF模式，两者的区别在这里主要体现在x_R表示式的不同， Two relay modes, AF mode and DF mode, the difference between the two is mainly reflected in the expression of x_R ,

(1)AF模式 (1) AF mode

$\begin{matrix} {x x}_{R R}^{AF AF} = = {ky ky}_{R R} = = \sqrt{\frac{{P P}_{R R}}{{| | {h h}_{SR SR} | |}^{22} {P P}_{S S} {d d}_{SR SR}^{- - α α} + + {| | {h h}_{CR CR} | |}^{22} {P P}_{C C} {d d}_{CR CR}^{- - α α} + + {N N}_{R R}}} {y the y}_{R R} \\ = = \sqrt{\frac{{P P}_{R R}}{{| | {h h}_{SR SR} | |}^{22} {P P}_{S S} {d d}_{SR SR}^{- - α α} + + {| | {h h}_{CR CR} | |}^{22} {P P}_{C C} {d d}_{CR CR}^{- - α α} + + {N N}_{R R}}} (({h h}_{SR SR} \sqrt{{P P}_{S S} {d d}_{SR SR}^{- - α α}} {x x}_{S S} + + {h h}_{CR CR} \sqrt{{P P}_{C C} {d d}_{CR CR}^{- - α α}} {x x}_{C C} + + {n no}_{R R})) \end{matrix} - - - - - - ((44))$

$k k = = \sqrt{\frac{{P P}_{R R}}{{| | {h h}_{SR SR} | |}^{22} {P P}_{S S} {d d}_{SR SR}^{- - α α} + + {| | {h h}_{CR CR} | |}^{22} {P P}_{C C} {d d}_{CR CR}^{- - α α} + + {N N}_{R R}}} = = \sqrt{\frac{{P P}_{R R}}{{G G}_{SR SR} {P P}_{S S} + + {G G}_{CR CR} {P P}_{C C} + + {N N}_{R R}}} - - - - - - ((55))$

其中，k代表放大增益，G_XY用来表示X-Y链路的信道增益，即

Normalizing x_R gives:

${x x}_{R R}^{AF AF} = = \sqrt{\frac{11}{{| | {h h}_{SR SR} | |}^{22} {P P}_{S S} {d d}_{SR SR}^{- - α α} + + {| | {h h}_{CR CR} | |}^{22} {P P}_{C C} {d d}_{CR CR}^{- - α α} + + {N N}_{R R}}} (({h h}_{SR SR} \sqrt{{P P}_{S S} {d d}_{SR SR}^{- - α α}} {x x}_{S S} + + {h h}_{CR CR} \sqrt{{P P}_{C C} {d d}_{CR CR}^{- - α α}} {x x}_{C C} + + {n no}_{R R})) - - - - - - ((66))$

$\begin{matrix} {y the y}_{D D.}^{AF AF} = = {kh kh}_{SR SR} \sqrt{{P P}_{S S} {d d}_{SR SR}^{- - α α}} {h h}_{RD RD} \sqrt{{d d}_{RD RD}^{- - α α}} {x x}_{S S} + + k k {h h}_{CR CR} \sqrt{{P P}_{C C} {d d}_{CR CR}^{- - α α}} {h h}_{RD RD} \sqrt{{d d}_{RD RD}^{- - α α}} {x x}_{C C} \\ + + k k {h h}_{RD RD} \sqrt{{d d}_{RD RD}^{- - α α}} {n no}_{R R} + + {h h}_{CD cd} \sqrt{{P P}_{C C} {d d}_{CD cd}^{- - α α}} {x x}_{C C} + + {n no}_{D D.} \end{matrix} - - - - - - ((77))$

$\begin{matrix} {γ γ}_{D D.}^{AF AF} = = \frac{{k k}^{22} {| | {h h}_{SR SR} | |}^{22} {d d}_{SR SR}^{- - α α} {| | {h h}_{RD RD} | |}^{22} {d d}_{RD RD}^{- - α α} {P P}_{S S}}{} \\ = = \frac{{k k}^{22} {G G}_{SR SR} {G G}_{RD RD} {P P}_{S S}}{{k k}^{22} {G G}_{CR CR} {G G}_{RD RD} {P P}_{C C} + + {k k}^{22} {G G}_{RD RD} {N N}_{R R} + + {G G}_{CD cd} {P P}_{C C} + + {N N}_{D D.}} \end{matrix} - - - - - - ((88))$

(2)DF模式 (2) DF mode

${y the y}_{D D.}^{DF DF} = = {h h}_{RD RD} \sqrt{{P P}_{R R} {d d}_{RD RD}^{- - α α}} {x x}_{S S} + + {h h}_{CD cd} \sqrt{{P P}_{C C} {d d}_{CD cd}^{- - α α}} {x x}_{C C} + + {n no}_{D D.} - - - - - - ((1010))$

${γ γ}_{R R}^{DF DF} = = \frac{{G G}_{SR SR} {P P}_{S S}}{{G G}_{CR CR} {P P}_{C C} + + {N N}_{R R}} - - - - - - ((1111))$

${γ γ}_{D D.}^{DF DF} = = \frac{{G G}_{RD RD} {P P}_{R R}}{{G G}_{CD cd} {P P}_{C C} + + {N N}_{D D.}} - - - - - - ((1212))$

对于D2D接收端的最终信号接收，我们采用基于最大比的分集合并(MSC)方式，如图3所示，在D处进行两个支路信号的最大比合并，并且分析可知最大比合并后的总的SINR为各支路SINR之和： For the final signal reception at the D2D receiving end, we adopt the method based on maximum ratio diversity combining (MSC). The SINR is the sum of the SINR of each branch:

${γ γ}_{D D.} = = {γ γ}_{D D.}^{direct directly} + + {γ γ}_{D D.}^{relay relay} - - - - - - ((1313))$

其中

表示直接链路的SINR，

表示中继链路的SINR： in

Indicates the SINR of the direct link,

Indicates the SINR of the trunk link:

${γ γ}_{D D.}^{direct directly} = = \frac{{| | {h h}_{SD SD} | |}^{22} {d d}_{SD SD}^{- - α α} {P P}_{S S}}{{| | {h h}_{CD cd} | |}^{22} {d d}_{CD cd}^{- - α α} {P P}_{C C} + + {N N}_{D D.}} = = \frac{{G G}_{SD SD} {P P}_{S S}}{{G G}_{CD cd} {P P}_{C C} + + {N N}_{D D.}} - - - - - - ((1414))$

上面的分析都是针对D2D系统，而对于蜂窝用户来说，蜂窝用户的SINR表达式为： The above analysis is all for the D2D system, but for the cellular users, the SINR expression of the cellular users is:

当S向R发送信号时 (16)

When S sends a signal to R (16)

当R向D发送信号时 (17)

When R sends a signal to D (17)

本发明一种基于终端直通中继通信的资源分配方法，其包括如下步骤： The present invention is a resource allocation method based on terminal direct relay communication, which includes the following steps:

步骤1:假设所有的蜂窝用户以固定功率P_CO进行通信，而所有的D2D用户的发送功率则固定为P_SO，所有的用户终端(UE)（包括活跃的、空闲的）蜂窝用户和D2D用户都一直与基站保持着连接并周期地与基站交换信息，另外，即使是正在进行D2D通信的用户，仍然要受基站控制； Step 1: Assuming that all cellular users communicate with a fixed power P_CO , while the transmission power of all D2D users is fixed at P_SO , all user terminals (UE) (including active and idle) cellular users and D2D users They are always connected with the base station and periodically exchange information with the base station. In addition, even users who are communicating with D2D are still subject to the control of the base station;

步骤2:上行蜂窝信道集合记为

，所有的无线资源分配和调度都由基站完成，基站维持着一个本小区资源分配实时状况的信道记录表，记作ChnTable，如表1所示，每一行都表示一个信道的资源使用情况，每行的第1格记录信道号n，第2格记录使用该信道的蜂窝用户C_UE_i(n)，第3格和第4格表示复用该信道的D2D用户对S_UE_j(n)、D_UE_j(n)，最后一格表示选取的移动中继I_UE_k(n)；若第2格为空表示该信道还未被蜂窝用户使用，其余格为空表示该信道没有被D2D用户复用。只要资源分配信息发生了变化，基站就会对ChnTable进行更新，一个信道内最多同时共存一个蜂窝用户和一对基于移动中继的D2D用户； Step 2: The set of uplink cellular channels is denoted as

, all wireless resource allocation and scheduling are done by the base station. The base station maintains a channel record table of the real-time resource allocation status of the cell, which is denoted as ChnTable. As shown in Table 1, each row represents the resource usage of a channel. The first grid of the row records the channel number n, the second grid records the cellular user C_UE_i(n) using this channel, the third and fourth grids indicate the D2D user pairs S_UE_j(n) and D_UE_j that multiplex this channel_(n) , the last grid represents the selected mobile relay I_UE_k(n) ; if the second grid is empty, it means that the channel has not been used by cellular users, and if the remaining grids are empty, it means that the channel has not been reused by D2D users. As long as the resource allocation information changes, the base station will update the ChnTable, and at most one cellular user and a pair of mobile relay-based D2D users can coexist in one channel at the same time;

步骤3:N个蜂窝用户C_UEs使用N个不同的上行蜂窝信道与BS进行通信，彼此之间没有干扰，资源分配工作由基站完成，并在ChnTable中记录更新； Step 3: N cellular users C_UEs use N different uplink cellular channels to communicate with the BS, there is no interference between each other, the resource allocation work is completed by the base station, and records are updated in the ChnTable;

步骤4:当有潜在的D2D通信可能时，D2D发送用户S_UE和D2D接收用户D_UE分别将它们的位置信息报告给基站，这一过程利用内置的全球定位系统(GPS)或者北斗卫星导航系统(BDS)模块来完成，并根据干扰模型来进行蜂窝用户和D2D用户的信号干扰噪声比(SINR)计算，由基站来决定是否准许进行D2D尝试，如果不合适的话，该对D2D用户会被基站拒绝并进入等待队列： Step 4: When potential D2D communication is possible, the D2D sending user S_UE and the D2D receiving user D_UE report their location information to the base station respectively. This process utilizes the built-in Global Positioning System (GPS) or Beidou Satellite Navigation System (BDS ) module, and calculate the signal-to-interference-noise ratio (SINR) of the cellular user and the D2D user according to the interference model, and the base station decides whether to allow the D2D attempt. If it is not appropriate, the pair of D2D users will be rejected by the base station and Enter the waiting queue:

如图4所示流程，首先由基站检查ChnTable，去除已经饱和即一个蜂窝用户和一对D2D用户共存的信道，剩余信道集合记为

，如果其不为空的话，分别对其中的各个信道进行进一步分析，看能否容纳下该对D2D用户使得该信道下所有用户都能满足通信要求，筛选出的合格信道即蜂窝用户SINR达到蜂窝用户的目标SINRγ_C-target，D2D用户SINR达到D2D用户的目标SINRγ_D-target，组成集合

，并根据改进的遗传算法从

中选出目标信道供该D2D对复用，遗传算法的进化目标选取的是最小化资源分配失败的概率； As shown in Figure 4, firstly, the base station checks the ChnTable, and removes channels that are saturated, that is, a cellular user and a pair of D2D users coexist, and the remaining channel set is denoted as

, if it is not empty, further analyze each of the channels to see if the pair of D2D users can be accommodated so that all users under the channel can meet the communication requirements. The target SINRγ_C-target of the user, and the SINR of the D2D user reaches the target SINRγ_D-target of the D2D user, forming a set

, and according to the improved genetic algorithm from

步骤5:进行移动中继的选择，以D2D用户对收发端连线为直径构成的圆形区域中总有若干空闲用户可供选取作为移动中继，从中为每对D2D用户选取m个空闲用户作为候选中继集合，一个空闲用户只会出现在一对D2D用户的候选中继集合中，同一时刻一个空闲用户只能为一对D2D用户服务，一对D2D用户只能选择一个空闲用户作为移动中继； Step 5: Select a mobile relay. There are always a number of idle users in the circular area formed by the diameter of the D2D user's connection to the transceiver end to be selected as a mobile relay, and m idle users are selected for each pair of D2D users. As a candidate relay set, an idle user will only appear in the candidate relay set of a pair of D2D users. At the same time, an idle user can only serve a pair of D2D users, and a pair of D2D users can only choose one idle user as the mobile relay;

依据D2D中继系统SINR最优原则，并根据干扰模型来进行蜂窝用户和D2D用户的SINR计算，从候选移动中继集合中选取出移动中继，使得该D2D中继通信可以获得最大的SINR； According to the optimal principle of D2D relay system SINR, and according to the interference model to calculate the SINR of the cellular user and the D2D user, select the mobile relay from the candidate mobile relay set, so that the D2D relay communication can obtain the maximum SINR;

若某个空闲用户电量不足或不适合用于移动中继，则从候选中继集合中去除该空闲用户； If an idle user has insufficient power or is not suitable for mobile relay, remove the idle user from the candidate relay set;

本发明中采用的遗传算法，具体流程如图5所示，先生成初代群体，对基因进行实数编码，再对每一代群体依次进行选择、交叉、变异操作，直至生成最终代（记为T）群体，包括以下具体步骤： The genetic algorithm used in the present invention, the specific process is shown in Figure 5, the first generation population is first generated, the genes are encoded with real numbers, and then each generation population is sequentially selected, crossed, and mutated until the final generation (denoted as T) is generated. groups, including the following specific steps:

步骤1:对基因进行实数编码；染色体长度为蜂窝信道数N，共含有N个基因，对基因进行实数编码，一个信道最多只能被一对D2D用户复用，一对D2D用户最多只能复用一个信道，某个基因值为0表示该信道没有被D2D用户复用，第i个基因的数值A_i表示第A_i对D2D用户复用的是第i个信道，因而基因型A＝A₁A₂…A_N所对应的表现型为x＝(x₁,x₂,…,x_N)^H；即： Step 1: Encode the genes with real numbers; the length of the chromosome is the number of cellular channels N, and there are N genes in total. The genes are encoded with real numbers. A channel can only be multiplexed by a pair of D2D users at most, and a pair of D2D users can only multiplex at most. Using a channel, a gene value of 0 indicates that the channel is not reused by D2D users, and the value A_i of the i-th gene indicates that the i_- th channel is multiplexed by the A i-th pair of D2D users, so genotype A=A₁ A₂ …A_N corresponds to x=(x₁ ,x₂ ,…,x_N )^H ; that is:

A_i＝x_i (18) A_i = x_i (18)

显然，A_i＝0表示该信道没有被任何一对D2D用户复用，而对于

, if A_i ≠0 and A_j ≠0, then A_i ≠A_j ;

步骤2:对当前代数的种群进行选择操作；由于进化目标选取的是最小化资源分配失败的概率，因此用D2D用户的资源分配成功率来评价个体的适应度，如下式，用于选择操作，其中

Indicates the i-th individual in the l-generation population;

(1)计算个体在下一代群体中的期望生存数目，用Q_i来表示，如下式： (1) Calculate the expected survival number of individuals in the next generation group, expressed by_Qi , as follows:

(2)确定各对应个体在下一代群体中的准生存数目，

向上取整，从而得到准下一代群体的个体

个； (2) Determine the quasi-survival number of each corresponding individual in the next generation group,

Round up to get the individuals of the quasi-next generation group

indivual;

(3)从准下一代群体中删去适应度最低的

individuals, so as to obtain Q selected individuals;

步骤3:对经过选择操作的种群进行行内交叉；采用行内交叉的操作，对经过选择操作的种群进行行内交叉，确保生成的新个体仍是有效的信道排列，能够作为一种信道分配方案； Step 3: Carry out intra-row crossover to the population through the selection operation; adopt the operation of intra-row crossover, carry out intra-row crossover to the population through the selection operation, ensure that the new individual generated is still an effective channel arrangement, and can be used as a channel allocation scheme;

步骤4:对经过交叉操作的种群进行均匀变异，确保能够生成新的信道分配方案； Step 4: uniformly mutate the population through the crossover operation to ensure that a new channel allocation scheme can be generated;

步骤5:重复步骤2-步骤4，直到进化到最终代数T，从中选取出适应度最高的个体作为资源分配方案。 Step 5: Repeat step 2-step 4 until it evolves to the final algebra T, and select the individual with the highest fitness as the resource allocation plan. the

Claims

1. the resource allocation methods based on the communication of terminal straightforward trunking, it is characterized in that: in single subdistrict, LTE-Advanced cellular network and D2D system form hybrid network, the duplex mode of cellular network is time division duplex, D2D utilizes the ascending resource of cellular communication in community with multiplexing pattern, comprise the process of choosing with the application of the renewal of mutual, the resource record table of base station, resource and distribution, via node; First according to poll criterion, guarantee that communication quality is prerequisite, system resource allocation is given and enlivened phone user, then according to maximum resource utilance criterion, utilize D2D user in genetic algorithm Lai Wei community to select most suitable resource to carry out multiplexing, target is that minimized resource distributes failed probability, is guaranteeing, on the basis of phone user's proper communication, more D2D user to be communicated; Comprise the following steps:

Step 1: suppose that all phone users are with constant power P_cOcommunicate, all D2D users' transmitted power is fixed as P_sO, all user terminal UE, comprise that active, idle phone user and D2D user are keeping being connected also periodically and base station exchange information with base station always, even carry out the user of D2D communication, controlled by base station;

Step 2: ascending honeycomb channel set is designated as

all allocation of radio resources and scheduling are all completed by base station, base station maintain the channel record table that this local resource distributes real time status, be denoted as ChnTable, for recording the service condition of each channel, as long as variation has occurred resource allocation information, base station will be upgraded it, and a phone user and a pair of D2D user based on mobile relay coexist at most in a channel simultaneously;

Step 3:N phone user C_UEs uses N different ascending honeycomb channel and base station to communicate, and do not disturb each other, and resource shares out the work and completed by base station, and record upgrades in ChnTable;

Step 4: in the time having potential D2D communication possibility, D2D send user S_UE and receive user D_UE respectively by their location information report to base station, this process utilizes built-in global position system GPS or Beidou satellite navigation system BDS module to complete, and according to interference model carry out phone user and D2D user Signal Interference and Noise Ratio SINR calculate, determine whether permitting carrying out D2D trial by base station, if inappropriate words, this can be refused and enter by base station waiting list to D2D user;

First check ChnTable by base station, removal is saturated is the channel that a phone user and a pair of D2D user coexist, and residue channel set is designated as

if it is not empty, respectively wherein each channel is further analyzed, sees that can hold lower this makes all users under this channel can meet communicating requirement to D2D user, the qualified channel filtering out is the target SINR γ that phone user SINR reaches phone user_c-target, D2D user SINR reaches D2D user's target SINR γ_d-target, composition set

, and according to improved genetic algorithm from

in select destination channel for this D2D to multiplexing, what the evolution target of genetic algorithm was chosen is that minimized resource distributes failed probability;

If do not find possible channels, this group of D2D user will be rejected and enter waiting list;

Step 5: carry out the selection of mobile relay, using D2D user, sending and receiving end line always there are in the border circular areas that diameter forms some free subscribers can supply to choose as mobile relay, therefrom choose m free subscriber as candidate relay set for every couple of D2D user, a free subscriber only there will be in a pair of D2D user's candidate relay set, free subscriber of synchronization can only be a pair of D2D user's service, and a pair of D2D user can only select a free subscriber as mobile relay; According to D2D relay system SINR principle of optimality, and carry out phone user and D2D user's SINR calculating according to interference model, from the set of candidate's mobile relay, select mobile relay, make this D2D trunking traffic can obtain maximum SINR; If certain free subscriber electric weight is not enough or be not suitable for mobile relay, from candidate relay set, remove this free subscriber;

Step 6: the mobile relay that utilizes channel that step 4 selects and step 5 to select, start to carry out D2D test communications, if actual test shows the interior cellular communication of this channel and D2D trunking traffic and can both be protected, allow to carry out the information recording of formal D2D trunking traffic and base station renewal ChnTable; Otherwise, return to step 4 and re-start channel selection, until by test;

Step 7: once there is user to complete the communication of oneself, or there is available channel resource, the information in ChnTable will be upgraded also immediately order successively and again the D2D user in waiting list be performed step to 4; In order to guarantee to continue reliable communication process, the communication that base station can be carried out the user who records in ChnTable in the cycle monitors.

Described genetic algorithm comprises the steps:

Step 1: gene is carried out to real coding; Chromosome length is that honeycomb channel is counted N, contain altogether N gene, gene is carried out to real coding, a channel at most can only be multiplexing by a pair of D2D user, a pair of D2D user at most can only a multiplexing channel, certain genic value is that this channel of 0 expression is not multiplexing by D2D user, the numerical value A of i gene_irepresent A_ito D2D user multiplexing be i channel, thereby genotype A=A₁a₂a_ncorresponding phenotype is x=(x₁, x₂..., x_n)^h; That is:

A_i＝x_i

Obviously, A_i=0 represents that this channel is not multiplexing by any a pair of D2D user, and forif, A_i≠ 0 and A_j≠ 0, there is A_i≠ A_j;

Step 2: the population of current algebraically is selected to operation; What choose due to evolution target is that minimized resource distributes failed probability, therefore evaluates individual fitness with D2D user's Resource alloc success probability, as shown in the formula, for selecting operation, wherein

represent that l is for i individuality in colony;

According to based on upper limit fixed pattern sampling really, reduce the possibility that causes degeneration because of the contingency of selection, guarantee to select the most excellent individuality; Colony's number represents with Q:

(1) calculate individual expectation existence number in colony of future generation, use Q_irepresent, as shown in the formula:

Q_{i} = Q \cdot F (A_{i}^{l}) / Σ_{i = 1}^{Q} F (A_{i}^{l})

(2) determine each corresponding individual standard existence number in colony of future generation,

round up, thereby obtain the individuality of accurate colony of future generation

individual;

(3) from standard colony of future generation, leave out fitness minimum

individuality, thus Q the individuality through selecting obtained;

Step 3: to go interior intersection through the population of selecting operation; Adopt the operation intersecting in row, to go interior intersection through the population of selecting operation, guarantee that the new individuality generating is still effective channel and arranges, can serve as a kind of channel assignment scheme;

First comply with crossover probability p_cto through selecting the colony after operation to select at random the individuality that will intersect, but no longer it is matched between two, but to each elect wait intersect individual, all by random two random number i, the j (i ≠ j), and exchange A that generates 1 to N_iand A_jnumerical value, if the new individuality producing meets feasible solution requirement, new individuality is retained, if do not meet value requirement, deletes this individuality, and continues to produce by interlace operation new individual until meet feasible solution requirement;

Step 4: the population through interlace operation is evenly made a variation, guarantee to generate new channel assignment scheme;

First comply with variation Probability p_mfrom the colony after intersecting, select at random the individuality that need to make a variation, the individuality that need to make a variation to each generates at random a random number k, to Gene A by 1 to N_kcarry out mutation operation, from 1 ..., N}-{A_i| i=1 ..., in this difference set of N}, select at random an element, assign it to A_kas sudden change, if this difference set is empty, be set to 0.Produce the individual parameter value requirement that meets solution space if new, new individual will reservation, if do not meet the requirement of solution space parameter value, delete this individuality, and continue to produce by variation new individual until meet parameter value requirement;

Step 5: repeating step 2-step 4, until evolve to final algebraically T, therefrom selects individuality that fitness is the highest as Resource Allocation Formula.

2. according to a kind of resource allocation methods based on the communication of terminal straightforward trunking described in claims 1, it is characterized in that, in described step 4 and step 5, carry out phone user and D2D user's SINR calculating according to interference model, comprise the steps:

Steps A: in D2D trunking traffic model, phone user C carries out uplink communication with base station B, S-R-D is a D2D relay communications system, and multiplexing is the Radio Resource that C takies, and S represents that D2D sends user, and R represents mobile relay, D represents that D2D receives user.The mixed signal that has R, D, B tri-places to receive is respectively:

y_{R} = h_{SR} \sqrt{P_{S} d_{SR}^{- α}} x_{S} + h_{CR} \sqrt{P_{C} d_{CR}^{- α}} x_{C} + n_{R}

y_{D} = h_{RD} \sqrt{P_{R} d_{RD}^{- α}} x_{R} + h_{CD} \sqrt{P_{C} d_{CD}^{- α}} x_{C} + n_{D}

X_s, x_r, x_cbe respectively corresponding transmitted signal, P_s, P_r, P_cfor corresponding transmitted power, n_r, n_d, n_brepresent corresponding additive white Gaussian noise AWGN, corresponding noise power is respectively N_r, N_d, N_b, α represents path loss index; In addition, h_xY, d_xYbe respectively used to represent channel coefficients, the communication distance of X-Y link.

Step B: two kinds of repeater modes, AF and DF pattern, both differences are mainly reflected in x here_rthe difference of expression, calculate D2D user's SINR according to different repeater modes:

Under AF pattern, the signal receiving is carried out amplification forwarding by R, do not carry out information decoding.

\begin{matrix} x_{R}^{AF} = {ky}_{R} = \sqrt{\frac{P_{R}}{{| h_{SR} |}^{2} P_{S} d_{SR}^{- α} + {| h_{CR} |}^{2} P_{C} d_{CR}^{- α} + N_{R}}} y_{R} \\ = \sqrt{\frac{P_{R}}{{| h_{SR} |}^{2} P_{S} d_{SR}^{- α} + {| h_{CR} |}^{2} P_{C} d_{CR}^{- α} + N_{R}}} (h_{SR} \sqrt{P_{S} d_{SR}^{- α}} x_{S} + h_{CR} \sqrt{P_{C} d_{CR}^{- α}} x_{C} + n_{R}) \end{matrix}

k = \sqrt{\frac{P_{R}}{{| h_{SR} |}^{2} P_{S} d_{SR}^{- α} + {| h_{CR} |}^{2} P_{C} d_{CR}^{- α} + N_{R}}} = \sqrt{\frac{P_{R}}{G_{SR} P_{S} + G_{CR} P_{C} + N_{R}}}

Wherein, k represents gain amplifier, G_xYbe used for representing the channel gain of X-Y link,

to x_rbe normalized and obtain:

x_{R}^{AF} = \sqrt{\frac{1}{{| h_{SR} |}^{2} P_{S} d_{SR}^{- α} + {| h_{CR} |}^{2} P_{C} d_{CR}^{- α} + N_{R}}} (h_{SR} \sqrt{P_{S} d_{SR}^{- α}} x_{S} + h_{CR} \sqrt{P_{C} d_{CR}^{- α}} x_{C} + n_{R})

The reception signal, the SINR expression formula that finally obtain R place are respectively:

\begin{matrix} y_{D}^{AF} = {kh}_{SR} \sqrt{P_{S} d_{SR}^{- α}} h_{RD} \sqrt{d_{RD}^{- α}} x_{S} + k h_{CR} \sqrt{P_{C} d_{CR}^{- α}} h_{RD} \sqrt{d_{RD}^{- α}} x_{C} \\ + k h_{RD} \sqrt{d_{RD}^{- α}} n_{R} + h_{CD} \sqrt{P_{C} d_{CD}^{- α}} x_{C} + n_{D} \end{matrix}

\begin{matrix} γ_{D}^{AF} = \frac{k^{2} {| h_{SR} |}^{2} d_{SR}^{- α} {| h_{RD} |}^{2} d_{RD}^{- α} P_{S}}{k^{2} {| h_{CR} |}^{2} d_{CR}^{- α} {| h_{RD} |}^{2} d_{RD}^{- α} P_{C} + k^{2} {| h_{RD} |}^{2} d_{RD}^{- α} N_{R} + {| h_{CD} |}^{2} d_{CD}^{- α} P_{C} + N_{D}} \\ = \frac{k^{2} G_{SR} G_{RD} P_{S}}{k^{2} G_{CR} G_{RD} P_{C} + k^{2} G_{RD} N_{R} + G_{CD} P_{C} + N_{D}} \end{matrix}

Under DF pattern, can R place relate to a problem that be correctly decoded:

Here suppose that R place always can be correctly decoded out x_sthereby D place receives signal, and R, the SINR separately of D place are:

y_{D}^{DF} = h_{RD} \sqrt{P_{R} d_{RD}^{- α}} x_{S} + h_{CD} \sqrt{P_{C} d_{CD}^{- α}} x_{C} + n_{D}

γ_{R}^{DF} = \frac{G_{SR} P_{S}}{G_{CR} P_{C} + N_{R}}

γ_{D}^{DF} = \frac{G_{RD} P_{R}}{G_{CD} P_{C} + N_{D}}

Step C: high specific merges MSC; Final signal for D2D receiving terminal receives, and adopts the diversity based on high specific to merge mode, and the high specific that carries out two tributary signals at D place merges, and the total SINR analyzing after known high specific merges is each branch road SINR sum:

γ_{D} = γ_{D}^{direct} + γ_{D}^{relay}

Wherein,

represent the SINR of direct link,represent the SINR of repeated link:

γ_{D}^{direct} = \frac{{| h_{SD} |}^{2} d_{SD}^{- α} P_{S}}{{| h_{CD} |}^{2} d_{CD}^{- α} P_{C} + N_{D}} = \frac{G_{SD} P_{S}}{G_{CD} P_{C} + N_{D}}

Step D: the SINR that calculates phone user; Analysis is above all for D2D system, and for phone user, phone user's SINR expression formula is:

when S is during to R transmitted signal;

when R is during to D transmitted signal.