In recent years, wireless sensor networks (WSNs) have emerged as a new wireless network technology for distributed computing and embedded systems. With the proliferation of wireless sensor networks, they are envisioned to be applicable to the wide app ...

In recent years, wireless sensor networks (WSNs) have emerged as a new wireless network technology for distributed computing and embedded systems. With the proliferation of wireless sensor networks, they are envisioned to be applicable to the wide application domains including ubiquitous computing as well as military, industry, environment, home, and office applications. Data aggregation is one of the most efficient approaches to achieve significant energy savings in the WSNs. However, data aggregation has a tradeoff relationship with time delay which is another important performance measure of network. Moreover, most of the data in real world generally have a time constraint to be delivered to the destination. Therefore, it is needed to deliver the data within the delay bound while achieving the energy savings via aggregation in the wireless sensor networks.


This research work proposes both centralized and distributed approaches to achieve the timeliness as well as the low energy consumption. In the centralized approach, which is called aggregation time control (ATC), the aggregation time of the sensor node is controlled in order to aggregate and deliver the data from source nodes to the sink within the delay bound. The sink node gives the command to control the aggregation time of the sensor nodes. A distributed algorithm called distributed aggregation time adaptation (DATA) is proposed to eliminate the control overhead by yielding the role of adjusting the aggregation time to each node. Moreover, MAC delay is compensated to improve timeliness. DATA determines the aggregation time bound which represents how long the data are allowed to be aggregated. Moreover, a protocol to dynamically adapt the aggregation time at each sensor node accordingly is also proposed.


The proposed algorithms are designed to mainly cope with the asynchronous data generation from all nodes, thus eliminating the necessity of time synchronization. In addition, They are designed to be used without location information. These features make this study have low overhead. The proposed algorithms use the number of children and the depth of the nodes information to increase the aggregation performance. The performance of the proposed algorithms is evaluated in terms of an aggregation ratio and a miss ratio via the ns-2 simulation based on real parameters of MICA motes. The simulation results demonstrate that ATC and DATA successfully deliver the time-constrained data within the delay bound with low energy consumption. This study is applicable to the wireless sensor networks which require both the low energy consumption and the time-constrained data delivery.

본 연구를 통해 Wireless Sensor Networks (WSN)의 데이터 aggregation 을 수행함에 있어서 시간 제약을 가진 데이터를 처리할 수 있게 되었다. Soft real-time property 를 가진 데이터들을 deadline 안에 sink에 전달함과 동시에 aggregation 을 통해 에너지 소모를 줄 ...

본 연구를 통해 Wireless Sensor Networks (WSN)의 데이터 aggregation 을 수행함에 있어서 시간 제약을 가진 데이터를 처리할 수 있게 되었다. Soft real-time property 를 가진 데이터들을 deadline 안에 sink에 전달함과 동시에 aggregation 을 통해 에너지 소모를 줄일 수 있게 되었다. 이러한 연구 성과를 통해 WSN의 활용 범위가 더욱 넓어질 수 있다. 예를 들어, Plant에 설치된 센서 노드들의 주기적인 측정 데이터가 정해진 시간안에 전달됨으로써 데이터의 가치가 높아질 수 있으며, 가정내에 설치된 센서 노드들의 모니터링 데이터가 사용자에게 정해진 시간안에 도달하게 함으로써 사용자 만족도를 향상시킬 수 있을 것이다. 주기적 데이터의 실시간 모니터링을 통해 시스템 제어 분야와 같은 타 분야와의 연동도 가능해 질 수 있다.