中文摘要：

精准描述无线传感器网络（WSN）中受干扰节点的性能,对拥塞控制和速率分配等无线协议的有效运作具有重要意义.研究表明物理模型（PRR-SINR模型）在现有干扰模型中准确率较高,为此,分别提出集中式和分布式算法用于建立WSN节点的PRR-SINR模型.集中式算法通过一个中心节点控制节点收发测量包,使每个节点可以进行逐步建模;分布式算法则依赖每个节点自主控制收发包状况进行建模.在含有17个TelosB节点的WSN中对算法进行性能评估,结果表明,2种算法均能在保证高精确度的情况下,快速、低开销地对网络中的所有节点建立PRR-SINR模型.

英文摘要：

Measuring the performance of those nodes in Wireless Sensor Network（WSN） suffering from interference is meaningful for protocols such as congestion control and link scheduling.Recent studies suggest that the physical model （PRR-SINR model） is significantly more accurate than existing interference models.This paper proposes a central algorithm and a distributed algorithm to build the PRR-SINR model for every node in a WSN respectively.The central algorithm uses a node to send commands to tell other nodes in the network when to receive/broadcast measurement packets.Thereby each node will build the PRR-SINR model according to the commands.In the distributed algorithm,however,each node builds the model all by itself.This paper evaluates the two algorithms in a network which is composed of 17 TelosB nodes.Experimental result shows that the models built by both of the proposed algorithms achieve high accuracy,while the overhead is significantly low.