Research On Resource Allocation For D2D Discovery

Device-to-Device(D2D)communication,which promises direct communication between devices without relaying via base stations,turns out to be a key technology for LTE-A,due to its potential in improving the spectrum efficiency and reducing energy consumption.In order to facilitate successful D2D communication,it is crucial to perform neighbor discovery such that all the potential neighboring devices can be efficiently detected.Specifically,resource allocation plays an essential role in D2D discovery procedure,since it directly affects system performance such as discovery range,energy consumption,discovery ratio and delay.To address the above issue,we explicitly analyze state of the art resource allocation strategies including random and energy detection-based resource allocation.In addition,we discuss their advantages and disadvantages in detail.For the random resource allocation strategy,we illustrate the relation between performance-related metrics(e.g.,average discovery probability and discovery delay)and system parameters(e.g.,the number of users,the size of resource pool and discovery period)based on a simplified analytic model.Numerous of simulation results validate the efficiency of the analytic model and reveal the poor adaptation capability of existing resource allocation strategies.To this end,we propose three different resource allocation schemes.On the one hand,transmission probability adaption scheme is capable of improving neighbor discovery probability as well as discovery delay under the constrained resource circumstance.On the other hand,delay-based sub-frame adaption scheme and delay-based joint sub-frame and discovery period adaption scheme can provision the delay-sensitive service or serve as an universal adaption strategy to some extent.It should be emphasized that the joint sub-frame and discovery period adaption scheme can maximize the resource utilization while guaranteeing delay performance.Both numerical analysis and simulation results highlight the correctness and efficiency of the proposed schemes.