https://www.um.edu.mt/library/oar/handle/123456789/2068 2026-10-07T14:14:34Z https://www.um.edu.mt/library/oar/handle/123456789/149587 Title: DPICache : a distributed program image cache for wireless sensor networks Authors: Ellul, Joshua; Martinez, Kirk Abstract: Wireless sensor networks, like other computing platforms, require software updates from time to time due to software bugs, new functionality, better understanding of the surrounding environment or new applications. Physically connecting to and reprogramming each node is usually not feasible and often impossible. A number of wireless code distribution and reprogramming techniques have been presented. In this paper we present a novel technique to efficiently update any nodes that were lost during programming periods. Our experiments demonstrate a 57% decrease of bytes sent over the air when updating nodes that previously missed updates. 2008-10-01T00:00:00Z https://www.um.edu.mt/library/oar/handle/123456789/149586 Title: A dynamic size distributed program image cache for wireless sensor networks Authors: Ellul, Joshua; Martinez, Kirk; De Roure, David Abstract: Reprogramming node software over-the-air is an essential requirement in many wireless sensor network applications due to the inaccessibility of the deployed sensor nodes. Transmitting whole software images consumes a high amount of energy in proportion to updates, especially when they are small in size. Incremental updates have addressed this; however introduce the potential of a sensor node becoming out of sync when it misses an update. In this paper we present a dynamic size distributed program image cache that provides increased efficiency in reprogramming out of sync nodes and multi-purpose wireless sensor networks. 2009-05-01T00:00:00Z https://www.um.edu.mt/library/oar/handle/123456789/149584 Title: Run-time compilation of bytecode in wireless sensor networks Authors: Ellul, Joshua; Martinez, Kirk Abstract: Recent work on virtual machines for wireless sensor networks has demonstrated the bene ts of using a Java programming paradigm for resource constrained sensor networks. Results have shown that a virtual machine approach greatly su ers from interpretation overheads. We present run-time compi- lation of bytecode which leverages from a compact platform independent bytecode application encoding as well as an ef- cient program execution platform by converting bytecode to native code in situ. 2010-04-01T00:00:00Z https://www.um.edu.mt/library/oar/handle/123456789/149582 Title: Field deployment of low power high performance nodes Authors: Martinez, Kirk; Basford, Philip; Ellul, Joshua; Clarke, Richard Abstract: When deploying a sensor network into a harsh environment the need for high levels of fault tolerance and maximising the usage of available resources become extremely important. This has been achieved by implementing a highly fault tolerant system based on our Gumsense boards. These combine an ARM-based Linux system with an MSP430 for sensing and power-control. It also allows for dynamic schedule modifications based on the available power and can be synchronised with other systems without relying on direct communication, autonomous behaviour in case of total communications failure is also supported. A deployment on Vatnajökull, the largest ice-cap in Europe, has provided a long-term test for the systems and revealed strengths and weaknesses in the design decisions. 2010-06-01T00:00:00Z