Mobile Data Gathering with Space-Division Multiple Access in Wireless Sensor Networks* Miao Zhao, Ming Ma and Yuanyuan Yang Department of Electrical and Computer Engineering, State University of New York, Stony Brook, NY 11794, USA Abstract—Recent years have witnessed a surge of interest in efficient data gathering schemes in wireless sensor networks (WSNs). In this paper, we ad- dress this important issue in WSNs by adopting mobility and space-division multiple access (SDMA) technique to optimize system performance. Specif- ically, a mobile data collector, for convenience, called SenCar in this paper, is deployed in a WSN. It works like a mobile base station and polls each sensor while traversing its transmission range. Each sensor directly sends data to the SenCar without any relay so that the lifetime of sensors can be prolonged. We also consider applying SDMA technique to data gather- ing by equipping the SenCar with two antennas. With SDMA, two distinct compatible sensors may successfully make concurrent data uploading to the SenCar. Intuitively, if the SenCar can always simultaneously communicate with two compatible sensors, data uploading time can be cut into half in the ideal case. We focus on the problem of minimizing the total time of a data gathering tour which consists of two parts: data uploading time and moving time. To better enjoy the benefit of SDMA, the SenCar may have to visit some specific locations where more sensors are compatible, which may ad- versely prolong the moving path. Hence, an optimum solution should be a tradeoff between the shortest moving path and full utilization of SDMA. We refer to this optimization problem as mobile data gathering problem with SDMA, or MDG-SDMA for short. We formalize the MDG-SDMA problem into an integer program (IP) and then propose three heuristic algorithms that provide practically good solutions to the problem. Our simulation re- sults demonstrate that the proposed algorithms can greatly reduce the total data gathering time compared to the non-SDMA algorithm with only mini- mum overhead. I. I NTRODUCTION Wireless sensor networks (WSNs), composed of densely- deployed, low-cost, low-power, multifunctional sensors, have emerged as a new information-gathering paradigm for taking spatial and temporal measurements of a given set of parame- ters, such as temperature, of a field [1], [2]. Sensors are usually random deployed over a field without a pre-configured infras- tructure. Each of these sensors has the capabilities to monitor the environment, collect data and route data back to the sink [3]. Typically, most energy of a sensor is consumed on two major tasks: sensing the field and uploading data to the sink. Energy consumption on sensing is relatively stable since it only depends on the sampling rate. On the other hand, the energy consumption on data uploading is non-uniform among sensors. It strongly depends on the network topology and the location of the destined data sink. Thus how to efficiently aggregate the information from the scattered sensors, generally referred to as data gathering, is an important and challenging issue as it largely determines the lifetime of the sensor network. Due to tremendous practical interests, much research effort has been devoted to efficient data gathering in WSNs and several schemes have been proposed, such as distributed data compres- sion [4], [5], efficient transmission schedule [6]-[8], hierarchical infrastructure [9]-[12] and data MULEs [13], [14]. Scaglione and Servetto [4] and Marco, et al. [5] considered joint data com- pression and routing in a sensor network. Data compression can effectively reduce the amount of raw data that need to be sent. * The research work was supported in part by the U.S. National Science Foun- dation under grant number ECS-0427345 and U.S. Army Research Office under grant number W911NF-04-1-0439. Jain, et al. [6] modeled the interference in a WSN by a con- flict graph, and determined an optimal schedule for maximizing the flow of information toward the destination. Duarte-Melo and Liu [7] studied the capacity of data gathering under the proto- col model. El Gamal [8] further studied a similar problem to that in [7] and investigated whether collaborative transmission schemes could improve throughput. In [9]-[12], network per- formance of a WSN was characterized by a hierarchical infras- tructure, in which sensors are organized into clusters and cluster heads take the responsibility of forwarding data to the outside data sink. The results show that the hierarchical infrastructure is an efficient way to handle the scaling issue in large-scale sen- sor networks. Finally, different from other routing schemes, data MULEs schemes use a special type of mobile nodes for facilitat- ing connectivity between static sensors. Data MULEs take the burden of data routing away from sensors, which may be desir- able when sensors have limited energy and storage resources. Although the above schemes can perform data gathering in WSNs, there still exist some inefficiencies in these schemes. Specifically, the non-uniformity of energy consumption among sensors exists in the first three types of schemes, while the data MULEs schemes may cause relatively long delay in data for- warding. In this paper, we will further improve the performance of data gathering in WSNs by considering two critical factors: mobility and space-division multiple access (SDMA) technique. To the best of our knowledge, this is the first work that introduces SDMA technique to data gathering and explores the utility of a joint design of mobility and SDMA technique in data gathering schemes. The mobility we refer to in this paper is to deploy a mobile collector in the sensing field [15]-[18], which collects data from sensors at some specific positions. We call such a data gathering scheme mobile data gathering (MDG). There are three major ad- vantages that make a mobile collector perfectly suitable to data gathering applications in WSNs. First, it radically solves the non-uniformity of energy consumption among sensors. In gen- eral, the closer a sensor is located to the sink, the faster its energy will be depleted due to more packet forwarding it has to per- form. In case of the failure or malfunctioning of a sensor around the sink, network connectivity and coverage may not be guar- anteed. Even in a WSN with a hierarchical architecture, cluster heads will inevitably consume more energy than other sensors. To avoid the problem of cluster heads failing faster than other nodes, sensors can become cluster heads rotationally [9]. How- ever, since every sensor may possibly become a cluster head, each of them has to be “powerful” enough to handle incoming and outgoing traffic, which will increase the cost of the entire network. Furthermore, it may incur high overhead due to the frequent information exchange among sensors. By introducing a mobile collector, it is possible for each sensor to send data di- rectly to the mobile collector without any relay when the mobile This full text paper was peer reviewed at the direction of IEEE Communications Society subject matter experts for publication in the IEEE INFOCOM 2008 proceedings. 978-1-4244-2026-1/08/$25.00 © 2008 IEEE 1957 Authorized licensed use limited to: SUNY AT STONY BROOK. Downloaded on November 13, 2008 at 21:43 from IEEE Xplore. Restrictions apply.