Abstract—For the sensor network to operate successfully, the active nodes should maintain both sensing coverage and network connectivity. Furthermore, scheduling sleep intervals plays critical role for energy efficiency of wireless sensor networks. Traditional methods for sensor scheduling use either sensing coverage or network connectivity, but rarely both. In this paper, we use random scheduling for sensing coverage and then turn on extra sensor nodes, if necessary, for network connectivity. Simulation results have demonstrated that the number of extra nodes that is on with upper bound of around 9%, is small compared to the total number of deployed sensor nodes. Thus energy consumption for switching on extra sensor node is small. Keywords—Wireless sensor networks, energy efficient network, performance analysis, network coverage. I. INTRODUCTION IRELESS Sensor Networks (WSNs) have attracted a great deal of research attention due to their wide range of applications, ranging from military application (for intrusion detection) to civilian application, such as battlefield surveillance, tracking, machine failure, diagnosis, biological detection, home security, smart spaces, managing inventory. Wireless Sensor Networks consists of a large number of tiny sensor nodes that are densely deployed inside the phenomenon or very close to it. Each sensor is composed of sensors, processors, memory and wireless transceivers. Due to their small dimension within several cube millimeters [3], they have very limited power supply. Energy efficiency is a critical concern in wireless sensor network, since a WSN is expected to operate for long time with sensor node’s limited power supply and because of large number of sensors or hostile environment, charging or changing the battery is impossible. In addition to energy efficiency, sensing coverage and network connectivity are critical requirements in sensor networks. Sensing coverage for sensing the area and detection of the events, that can be considered as the measure of quality of service of sensor network [13]. The unit area is covered if any point in that area is within sensing range of an active node. The network is connected if any active node can communicate with any other active node. With connectivity information collected by sensing coverage can be sent to sink M. Mahdavi (phone: 60-89216837; e-mail: mina@vlsi.eng.ukm.my), M. Ismail (e-mail: mbi@eng.ukm.my), K. Jumari (e-mail: kbj@eng.ukm.my) and Z. M. Hanapi (e-mail: zurina@vlsi.eng.ukm.my) are with the Electrical, Electronic & systems Engineering Department, University Kebangsaan Malaysia, 43600 UKM Bangi Selangor, Malaysia. or base station. As the number of sensor nodes in sensor network is more than what is required by scheduling nodes to sleep and tuning off redundant sensor nodes can be achieved both coverage and energy efficiency at the same time. Researches [15], [4], [5], [11], [8], [10], [13], [9] are existing coverage- preserving scheduling scheme while GAF [14], SPAN [2], ASCENT [1] and STEM [7] are topology management protocols that maintain the network connectivity. This work deals with joint problem of sensing coverage and network connectivity without certain constrains such as grids or relation between the radio range and the sensing range. In addition with this method, each active sensor node knows at least one path to the sink node and so at the same time routing problem is solved and no additional routing protocols are needed. II. ALGORITHM DESIGN The work that has been done by [6] is taken as reference. First randomize scheduling algorithm [12] for sensing coverage has been designed that does not assume the availability of any location or directional information. It is a purely distributed algorithm thus scalable for large networks. Assume that sensor nodes constitute a set S. Given a number k, each sensor node randomly joins one of the k disjoint subsets of set S. once the k subset are determined they work alternatively. At any given time, there is only one subset working, and all the sensor node belonging to this subset will turn on. The intuition is that when the network is sufficiently dense, each subset alone will cover most part of the field. Fig. 1 shows an example. if there are eight sensor nodes (with IDs 0, 1, ..7) randomly deployed in a rectangular area. Let say there are two subsets S 0 and S 1 (k=2). Each sensor randomly select 0 or 1 and join one of the corresponding subsets S 0 or S 1 . Assume that sensor nodes 0, 3, 4, 6 select number 0 and thus join subset S 0 , and sensor nodes 1, 2, 5, 7 select number 1 and join subset S 1 . Then subset S 0 and S 1 work alternatively means that when sensor nodes 0, 3, 4, 6 (solid circles) are active, sensor nodes 1, 2, 5, 7 (dashed circles), fall asleep and vise versa. Performance of a Connected Random Covered Energy Efficient Wireless Sensor Network M. Mahdavi, M. Ismail, K. Jumari, and Z. M. Hanapi W International Journal of Electrical and Computer Engineering 5:4 2010 216