Scan-Based Movement-Assisted Sensor Deployment Methods in Wireless Sensor Networks Shuhui Yang, Minglu Li, and Jie Wu Abstract—The efficiency of sensor networks depends on the coverage of the monitoring area. Although, in general, a sufficient number of sensors are used to ensure a certain degree of redundancy in coverage, a good sensor deployment is still necessary to balance the workload of sensors. In a sensor network with locomotion facilities, sensors can move around to self-deploy. The movement-assisted sensor deployment deals with moving sensors from an initial unbalanced state to a balanced state. Therefore, various optimization problems can be defined to minimize different parameters, including total moving distance, total number of moves, communication/computation cost, and convergence rate. In this paper, we first propose a Hungarian-algorithm-based optimal solution, which is centralized. Then, a localized Scan-based Movement-Assisted sensoR deploymenT method (SMART) and its several variations that use scan and dimension exchange to achieve a balanced state are proposed. An extended SMART is developed to address a unique problem called communication holes in sensor networks. Extensive simulations have been done to verify the effectiveness of the proposed scheme. Index Terms—Dimension exchange, Hungarian method, load balance, movement-assisted, scan, sensor deployment, wireless sensor networks. Ç 1 INTRODUCTION W IRELESS sensor networks (WSNs) [1], [2] combine processing, sensing, and communications to form a distributed system capable of self-organizing, self-regulat- ing, and self-repairing. The application of WSNs ranges from environmental monitoring to surveillance to coordi- nated target detection. The efficiency of a sensor network depends on the coverage of the monitoring area. Although, in general, a sufficient number of sensors are used to ensure a certain degree of redundancy in coverage so that sensors can rotate between active and sleep modes, a good sensor deployment is still necessary to balance the workload of sensors. Mobile sensors [3] can be exploited to provide a redistribution. After an initial random deployment of sensors in the field, movement-assisted sensor deployment [4] can be applied, which uses a potential-field-based approach to move existing sensors by treating sensors as virtual particles subject to virtual forces. Basically, movement-assisted sensor deployment deals with moving sensors from an initial unbalanced state to a balanced state. Therefore, various optimization problems can be defined to minimize different parameters, including total moving distance, total number of moves, communication/computation cost, and convergence rate. More recently, some extended virtual force methods, such as those in [5] and [6], which are based on disk packing theory [7] and the virtual force field concept from robotics [8], are proposed. These methods simulate the attractive and repulsive forces between particles. Sensors in a relatively dense region will explode slowly according to each other’s repulsive force and head toward a sparse region. In this way, the whole monitoring area can achieve an even distribution of sensors. However, these methods may have long deployment times since sensors move independently, and they may even fail if all the sensors can achieve force balance but not load balance. We assume that sensors are deployed randomly into the square monitoring area without consideration of any physical obstacles. Then, if we partition the monitoring area into many small regions and use the number of sensors in a region as its load, the sensor deployment problem can be viewed as a load balance problem in traditional parallel processing, where each region corresponds to a processor and the number of sensors in a region corresponds to the load. The sensor deployment resembles the traditional load balance issue in parallel processing, with several key differences: . Different objectives. In traditional load balancing, the total moving distance rather than the number of moves is important whereas, in sensor networks, the number of moves is also important because of a relatively heavy energy consumption to start or stop a move. IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS, VOL. 18, NO. 7, JULY 2007 1 . S. Yang and J. Wu are with the Department of Computer Science and Engineering, Florida Atlantic University, 777 Glades Road, Boca Raton, FL 33431. E-mail: {syang1, jie}@cse.fau.edu. . M. Li is with the Department of Computer Science and Engineering, Shanghai Jiao Tong University, 1954 Huashan Road, Shanghai, PR China. E-mail: li-ml@cs.stju.edu.cn. Manuscript received 1 June 2006; revised 28 Sept. 2006; accepted 2 Oct. 2006; published online 9 Jan. 2007. Recommended for acceptance by S. Olariu. For information on obtaining reprints of this article, please send e-mail to: tpds@computer.org, and reference IEEECS Log Number TPDS-0140-0606. Digital Object Identifier no. 10.1109/TPDS.2007.1048. 1045-9219/07/$25.00 ß 2007 IEEE Published by the IEEE Computer Society