No abstract

Cyber Physical Systems are distributed systemsof-systems that integrate sensing, processing, networking and actuation. Aggregating physical data over space and in time emerges as an intrinsic part of data acquisition, and is critical for dependable decision making under performance and resource constraints. This paper presents a Linear Programming-based method for optimizing the aggregation of data sampled from geographically-distributed areas while satisfy timing, precision, and resource constraints. The paper presents experimental results for data aggregation, including a case study on gas detection using a network of sensors. I. INTRODUCTIONCyber-Physical Systems (CPS) are expected to provide ubiquitous integration of data acquisition, processing and control to provide dependable decision making in dynamic conditions [8], [11]. Data acquisition is often through networked embedded nodes that can sense over distributed areas [3]. CPS are essential for next generation applications, including future transportation systems, intelligent power grid, health care, homeland security, and many more [8], [13].CPS originate new challenges related to effective decision making in dynamic conditions. This includes having accurate insight into the physical data over distributed space and in time [13]. This is hard to offer if data has to be acquired under timing constraints through resource-limited embedded nodes that communicate over slow communication links. Data aggregation for networked embedded nodes, the topic of this paper, helps addressing these challenges. Aggregation produces more compact data descriptions [4], which simplify decision making algorithms and aid inferring properties of behavior and performance (e.g., stability and latency). It also helps compensating for noisy or missing data due to unreliable data sensing and tight performance and resource constraints. For example, the impact of slow data communication can be mitigated by reducing the transmission volume through predictions using aggregated data models.Data aggregation for CPS differs in several ways from approaches proposed for wireless sensor networks (WSNs). The main goal of data aggregation for WSN is improving the network performance (bandwidth, throughput, and energy consumption) [5], [6], [7], [9]. A feedback-based aggregation scheme is discussed in [5]. The scheme adapts to changing traffic conditions and time requirements. Tiny Aggregation