When an application on a wireless sensor network (WSN) exhibits poor post-deployment performance, users (programmers and administrators) usually do not know which nodes in the network, nor which pieces of debugging information are relevant to the determination of causes of the performance problem. Blind logging and collection of information on all nodes and/or links for the purpose of debugging, however, are energy expensive and sometimes ineffective. To address this problem, we need algorithms and tools that help users pinpoint the causes of application performance problems, and then provide useful hints for fixing them or further examination. To meet this need, we propose a data-centric approach called post-deployment performance debugging (PD2). PD2 focuses on the data flows that an application generates, and relates poor application performance to significant data losses or latencies of some data flows (problematic data flows) as they go through the software modules on individual nodes and through the network. PD2 derives a few inference rules based on the data dependencies between different software modules, as well as between different nodes, and use them to trace back in each problematic flow. Then, PD2 turns on the performance monitoring of, and collects debugging information from, only those modules and nodes that the problematic flows go through. Finally, PD2 provides the debugging information to help users isolate the causes of poor performance. We have implemented PD2 on TinyOS and evaluated it on a real WSN testbed. Our experimental results show that PD2 helps users quickly locate the possible sources of problems, such as code bugs, weak links, and radio interference. Depending on the hop-count between the source of the problem and the data sink, PD2's energy consumption (communication overhead) is shown to be only 5-10% of that of collecting debugging information from all nodes.

We proposeOpportunisticDataAggregation(OPAG) that incurs no computation error and tolerates moderate message losses in wireless sensor networks (WSNs). OPAG performs in-network data aggregation in two layers: (1) at the data-aggregation layer, aggregation results are computed accurately; and (2) at the data-routing layer,a WSN node may send intermediate/partial results to its aggregation node using multi-path routing in order to tolerate message losses.By space-multiplexing messages (i.e., padding multiple partial results or sensor readings in a single message), OPAG opportunistically exploits a multi-path routing scheme, which is more energy-efficient than retransmission. This is based on a key observation that, when sending a message, the radio may consume much more energy in idle listening during the backoff period and the time to wait for its acknowledgment than transmitting the data bits.We implemented OPAG on TinyOS 2.x and TMote Sky Mote, and evaluated its performance on the Motelab Testbed at Harvard University.When a network is relatively well connected, OPAG can reduce the energy consumption by 33% at the expense of slightly higher relative errors, compared to TAG with reliable transmission.Compared to synopsis diffusion, OPAG can reduce the aggregation errors by 50% while consuming roughly the same amount of energy.

Location services are essential to many applications running on a hybrid of wirelessly-networked mobile actors and static sensors, such as surveillance systems and the pursuer and evader game (PEG). To our best knowledge, there has been no previous location service protocol for wireless sensor networks. A number of location service protocols have been proposed for mobile ad hoc networks, but they are not applicable to sensor networks due to the usually large per-hop latency between sensors. This paper presents a distributed location service protocol (DLSP) for wireless sensor networks. Using a rigorous analysis of DLSP, we derive the condition for achieving a high packet-delivery ratio, and show how to configure the protocol parameters to ensure the scalability of DLSP. We find that DLSP is scalable if the mobile's speed is below a certain fraction of the packet-transmission speed, which depends on a movement threshold. For example, if the movement threshold for the location servers at the lowest level equals the radio range, the speed limit is one-tenth of the packet-transmission speed. The mobile's theoretical speed limit is one-fifth of the packet-transmission speed, beyond which DLSP cannot scale regardless of the movement threshold. Because of the high location-update overhead of DLSP, we propose an optimization, DLSP-SN, which can reduce the overhead by over 70%, while achieving high packet-delivery ratios. However, due to the griding effect, the packet's path length of DLSP-SN may be longer than that of DLSP, incurring higher data-delivery cost