International Journal of Information, Control and Computer Sciences ISSN: 2517-9942 Vol:3, No:1, 2009 62 Tree-on-DAG for Data Aggregation in Sensor Networks Prakash G L, Thejaswini M, S H Manjula, K R Venugopal, L M Patnaik Abstract—Computing and maintaining network structures for ef- ficient data aggregation incurs high overhead for dynamic events where the set of nodes sensing an event changes with time. Moreover, structured approaches are sensitive to the waiting time that is used by nodes to wait for packets from their children before forwarding the packet to the sink. An optimal routing and data aggregation scheme for wireless sensor networks is proposed in this paper. We propose Tree on DAG (ToD), a semistructured approach that uses Dynamic Forwarding on an implicitly constructed structure composed of multiple shortest path trees to support network scalability. The key principle behind ToD is that adjacent nodes in a graph will have low stretch in one of these trees in ToD, thus resulting in early aggregation of packets. Based on simulations on a 2,000-node Mica2- based network, we conclude that efficient aggregation in large-scale networks can be achieved by our semistructured approach. Keywords—Aggregation, Packet Merging, Query Processing. I. I NTRODUCTION T HE main operation of a wireless sensor network (WSN) is to monitor the physical environment, process the sensed information, and deliver the results to some specic sink nodes. Sensor nodes are normally powered by batteries with limited energy resource. Therefore, the primary challenge for this energy-constrained system is to design energy-efcient protocols to maximize the lifetime of the network [1]. Since radio transmission is the primary source of power consumption, the design of communication protocols for topology management, transmission power control, and energy-efcient routing. The basic idea is to route the packet through the minimum energy paths so as to minimize the overall energy consumption for delivering the packet from the source to the destination. The drawback of this approach is that it tends to overwhelm the nodes on the minimum energy path, which is undesirable for sensor networks since all sensor nodes are collaborating for a common mission and the duties of failed nodes may not be taken by other nodes. forwarding unaggregated packets increases with the scale of the network. To benefit from the strengths of structured and structureless approaches, we propose a semistructured approach. In order to use the sensor network efciently, the data these sensors produce must be properly accessed and Prakash G L, Thejaswini M, S H Manjula and K R Venugopal are with the Department of Computer Science and Engineering, University Visvesvaraya College of Engineering, Bangalore University, Bangalore 560 001, e-mail: glprakash@yahoo.com. L M Patnaik is a Vice Chancellor, Defence Institute of Advanced Technol- ogy(Deemed University), Pune, India. eventually propagated to the end user. One way this is done is through the organization of the sensor nodes into a network, which is treated and queried as a distributed sensor database. Such a sensor network allows for paths to be created from the root of the network to each and every sensor node in the network. However, sensor nodes are limited devices in terms of energy usage, node failures and lossy communication. In order to alleviate these limitations, especially of power constraints and fault tolerance, several schemes have been developed. One scheme that has shown much promise is to organize the nodes into a tree and synchronize the sending and receiving of packets from children to parents in that tree. Doing this allows for in-network aggregation to occur, which has been shown to lower the amount of energy used in sensor networks, thus extending their lifetime and usefulness. While query propagation and tree organization techniques have been shown to lower energy usage, they suffer in several respects, most importantly in the ability to deal with node crashes and failures. When one node fails all the nodes in its subtree will also be cut off from the network (due to the parent/child relationship) until the network can be reorganized. The problem is even more evident for nodes close to the root of the network, where a single node failure could cause a substantial portion of the tree to get isolated, dramatically decreasing the effectiveness of the data produced for the end user. challenges : The main challenge in designing such a protocol is to determine the packet forwarding strategy in absence of a preconstructed global structure to achieve early aggregation. Our approach uses a structureless technique locally, followed by Dynamic Forwarding on Tree on DAG (ToD), an implicitly constructed packet forwarding structure to support network scalability. After performing local aggregation, nodes dynamically decide the forwarding tree based on the location of the sources. The key principle behind ToD is that adjacent nodes in a graph will have low stretch in at least one of these trees in ToD, thus resulting in early aggregation of packets. contributions : This paper makes the following contribu- tions: • We propose an efficient and scalable data aggregation mechanism that can achieve early aggregation without