Abstract: This paper firstly transforms the conventional RSSI positioning model into a non-constrained expected value programming model. Secondly, one such model is solved by developing a novel fruit fly optimization algorithm in stochastic environments in order to seek the location of the unknown node. In this algorithm, the current fruit fly population is divided into sub-populations by arc grouping, and subsequently a hybrid mutation strategy is implemented to find the optimal solution. Comparative numerical experiments have validated that the algorithm, with high-efficient convergence and high positioning accuracy, is feasible in solving engineering RSSI positioning.

Key words: Received Signal Strength Indicator（RSSI）, expected value programming, fruit fly optimization, arc grouping, hybrid mutation

摘要： 针对未知节点的定位过度依赖于接收信号强度指示（Received Signal Strength Indicator，RSSI）物理测量的精度问题，将传统RSSI定位模型转化为非约束期望值规划模型，进而设计随机环境下的新型果蝇优化算法寻找未知节点的位置。该算法利用弧形分组将果蝇群均衡划分为子群，对果蝇个体实施混合变异，加速寻优进程，提高收敛速度和寻优精度。比较性的数值实验显示，该算法的收敛速度快，对未知节点的定位精度高，其应用于RSSI定位问题是可行的。

关键词: 接收信号强度指示, 期望值规划, 果蝇优化, 弧形分组, 混合变异