Academic Journals Database
Disseminating quality controlled scientific knowledge

Guest Editorial

Author(s): Kun Mean HOU

Journal: Journal of Networks
ISSN 1796-2056

Volume: 4;
Issue: 6;
Start page: 379;
Date: 2009;
Original page

Keywords: Special Issue | Wireless Sensor Networks

Wireless sensor network (WSN) is an emergent multi-disciplinary science, and it may be considered as the foundation of pervasive computing, mobile computing and wearable computing. WSN is a very active and competitive research area due to its diverse and unlimited potential applications: air, underground and underwater. In spite of its young age, economic impact of WSN is important, for examples the industrial control segment market will be worth $5.3B by 2010 and the smart home market will be worth $2.8 billion worldwide by 2012 (Source: Stamatis Karnouskos, EU-US 08 Workshop). WSN is a set of wireless nodes. On one hand, each wireless node (WN) has similar hardware and software functionalities as a PC: CPU, memory, operating system, and communication protocol to fulfill a specific task. On the other hand, a WN has a limited power supply (embedded battery) and consumes approximately 1 million less power (~100µW instead of ~100W) than a PC. Due to resources constraints: energy consumption and form factor, the approaches applied in general purpose computer systems are not adapted to the requirements of WSN.  When it comes to the design of energy efficient oriented hardware and software components of WSN, cross-layering optimization approaches are generally adopted such as application-specific unified hardware and software by taking into account the following criteria: trade-off between complexity, efficiency and resource consumption, and application context (context-aware) etc. Currently two main hardware development and design trends are carried out to implement the WN: Commercial Off-The-Shelf ‘COTS’ and System on Chip ‘SoC’. The first and second generations of WN were designed by using low power 8-bit or 16-bit microcontroller processor core, Bluetooth and non standard wireless access medium (MICA Mote). The current trend of WN design such as Tmotesky, iMote and LiveNode are based on low power 16-bit or 32-bit RISC microcontroller, and full compliance IEEE802.14.5 standard. However the ultimate goal of all the researchers in the world is the implementation of long life, low cost and invisible WN integrated and embedded into environment. Three key technologies make possible to achieve this objective: MEMS ‘MicroElectroMEchanical systems’, UWB ’Ultra-Wide Band’ and low power CMOS technology. Different WN prototypes are realized by Intel (iMOTE2), University of Michigan (MOTE: Michigan Uni Prototype) and University of Berkeley (Pico-Mote). WN hardware seems easier to solve than embedded software for diverse WSN applications. The main questions which are related to WSN basic software design is how to keep modularity, high level abstraction and reliability to enable to implement complex massively distributed WSNs to meet resource constraint requirements. Real-time operating system (RTOS) plays a key role to support high level abstraction and distributed collaborative processing. Currently four categories of WSN’s RTOS are developed: Event driven (TinyOS), Multitask (RETOS, tKernel, NutOS, MANTIS), Data-Centric (AmbientRT), and Hybrid (Contiki, LIMOS). Note that TinyOS is very popular but it not adapted to complex hard real-time application. The development challenges of the WSN RTOS are energy-efficiency (context aware, configurable, small footprint), robustness, fault tolerance, support hard real-time constraint, and support component based model (high level of abstraction to ease the integration of high level SW such as protocol, middleware, application, and simulator). Furthermore, for WSN applications, message sending is energy consuming. Thus it is important to implement embedded energy efficient wireless routing protocol to increase WSN lifetime. It is clear that general purpose MANET routing protocol such as AODV (active), OLSR (proactive) and ZBR (hybrid) etc. are not suitable for WSN due to resource constraints. For example optimal routing path is well adapted to general purpose MANET but not suitable for WSN because the repetitive use of the same path will exhaust the battery of WNs belonging to the optimal routing path (black hole). Many WSN dedicated protocols are implemented (spin, cougar, gear, leach, speed …) but it is currently very difficult to have a clear idea concerning their performance (energy consumption, scalability, connectivity, lifetime …) because of the lack of large scale WSN real world experimentation results and because the simulation model does not reflect the real-world ones (physical layer). Note that, there is no standard scenario and the application program (with a known number of WNs) enables to evaluate rationally the performance of wireless routing protocols. In addition routing protocol relies on the WSN topology. On one hand, an optimal WSN topology facilitates the implementation of routing and administration protocols. On the other hand, the deployment of large scale WS nodes in a large area is random and its topology is a priori unknown. Then, it is important to investigate the auto-configuration algorithms to increase the efficiency of routing and administration protocols. However the frontier between the administration protocol and the routing protocol is not as clearly defined as in a classical network (e.g. TCP/IP and SNMP) due to cross-layering approach. Moreover WSN security, reliability, and fault tolerance are still an open problem. In this special issue 5 papers are selected among 40 submitted papers for the NTMS workshop on wireless sensor network: theory and practice, held at Tangier at Morocco in 2008, the rest of the papers are selected from an open call for paper. WSN is a multi-disciplinary science. It impossible to present all its topics but this special issue addresses most of the WSN embedded software problems dealing with real-world applications (EU NeT-ADDED FP6 project, French ANR research project and industrial projects). The first paper investigates the operating system dedicated to WSN and proposes a new native hybrid real-time operating system, named HEROS ‘Hybrid Embedded Real-time Operating System’, which may be configured to run in different modes: event-driven, multitask or event-driven and multitask to adapt to diverse domains of WSN applications. Furthermore, to ease distributed cooperative application, HEROS adopts LINDA concept by providing a simplified tuple space and a lightweight IN/OUT primitive-pair to achieve system communication & synchronization. The second paper describes a new MANET protocol dedicated to intelligent transport system, named CIVIC (Communication Inter Véhicule Intelligente et Coopérative). The CIVIC protocol is a location-based routing protocol supporting infrastructure and ad-hoc networks. CIVIC is a hybrid protocol including reactive and proactive routing processes. The third paper presents an autonomous network structure including its target-oriented routing algorithm named SCAR ‘Sequential Coordinate Routing Algorithm’. Besides the target-oriented property as a main feature of the SCAR, based on the mathematical claim and its proof, it is shown how the minimum energy consumption is taken into consideration. The fourth paper investigates the QoS of routing protocol to meet the requirement of real-time monitoring applications. The cross-layering techniques from MAC to application layers are adopted to improve the performance of the whole network. The authors show that the proposed cross layering protocols enable to reduce the delay and minimize energy consumption. The fifth paper presents a new auto-configuration algorithm based on graph optimization techniques to minimize WSN energy consumption by taking into-account the robustness and the network coverage. The key issue of this problem is the energy-efficient topology design. In this work, four heuristic strategies are proposed for designing WSN topologies. The sixth paper proposes a density control algorithm for wireless sensor networks to keep as few sensors as possible in active state to achieve a connected coverage of a specific area of interest. Inactive sensors can turn off sensing modules to save energy. Unlike other algorithms, the proposed one does not rely on position information or ranging information of sensors. Providing a powerful synchronization system is one of the most important goals to be pursued if an efficient utilization of sensor networks has to be addressed. The seventh paper proposes a novel synchronization system based on Kohonens Self Organizing Maps (SOMs), able to provide some Artificial Intelligence features to sensor nodes. The simulation results demonstrate the effectiveness of this algorithm, and present some important features such as the properties that the synchronization error does not increase when the number of nodes grows and that the percentage of needed anchors nodes is really very low. The eighth paper presents an approach of component software reliability analysis which includes the benefits of both time domain, and structure based approaches. To overcome the deficiency of existing NHPP techniques that fall short of addressing repair and internal system structures simultaneously, a technique of transformation of testing data is proposed to improve prediction reliability. This new approach is evaluated and validated through an example to show its effectiveness. I hope that this special issue, which deals with basic embedded WSN software: operating system, routing protocol, auto-configuration, synchronization and component software reliability analysis will contribute significantly to advance the WSN research. Finally I would like to thank all the authors for their interesting contributions and the quality of their work, as well as all the Technical Program Committee members who also contributed significantly to improve the quality of this special issue. 
Save time & money - Smart Internet Solutions      Why do you need a reservation system?