Fast growth in the number of vehicles on roads is the prime source of ambient air pollution in big cities of Pakistan. A survey conducted for Lahore, the second largest city in Pakistan, under the United Nations office for coordination of humanitarian affairs has revealed a surge of 400% in the number of vehicles over last couple of decades. The situation is further aggravated as there are no plans to limit the number of new vehicle registrations in the metropolitan areas of the country. Any measures to address ambient air pollution related issues will require an accurate temporal as well as spatial estimate of pollutants in the air. The conventional solutions for air quality assessment in wide use today are highly expansive, allowing only few installations of the monitoring stations resulting in poor spatial resolution. In addition, monitoring stations perform data aggregation, providing only slow scale variations of pollutant concentrations. But total change in air pollutants round the clock providing an insight about fast scale variations is equally important as well. The issue of poor spatial and temporal resolution is addressed partially using wireless sensor networks (WSNs) for air quality monitoring, which also reduces cost of overall system. But air quality monitoring solutions based on WSNs have their own limitations. For instance, data transportation from a sensor node to processing center can involve an arbitrary number of hops for data relaying, resulting in effective data throughput reduction with every additional hop. In addition, WSNs for air quality monitoring are also vulnerable to malicious data transmissions from rouge nodes. This necessitates the design and development of secure data transmission protocols robust to hostile adversary attacks in WSNs.

The proposed research addresses limitations of existing solutions by developing an integrated wireless sensor and vehicular network (WSVN). The proposed cost effective solution can achieve fast and slow scale air quality monitoring at the expense of slight accuracy degradation. The integrated WSVN addresses data transportation issue by limiting the number of hops required from sensor node to data processing center, to only two for best case scenario (i.e from sensor node to vehicular node and from vehicular node to gateway with wired connection between gateway and the data processing center). And under the worst case, only those sensor nodes which can not send their data directly to the vehicular nodes are allowed to route their data through intermediate sensor nodes. The benefit due to the reduction in number of hops will be two folds, 1) providing improved data throughput rates, and 2) an avenue to reduce the complexity of security enabled protocols.

The proposed research will lead to system architecture for secure and reliable air quality data acquisition, processing and visualization suggesting for countermeasures on both short term and long term basis. Particularly, the proposed approach will employ advanced stochastic modeling and control techniques providing optimal policies for resource management at the node as well as network level. Specifically, stochastic control solution employing Markov Decision Processes (MDPs) will be developed for this purpose. The optimal policies based on MDPs will achieve a balance between competition among the nodes for preserving their resources and cooperation to achieve the global objective. To ensure data integrity along the path from sensor nodes to processing center, we will develop secure data transfer protocols which will provide fast, robust and reliable node authentication
for data exchange while reducing the implementation complexity.

This research effort will have a positive impact on the life of ordinary people of Pakistan by first quantifying the concentration of pollutants in the ambient air and then suggesting countermeasures to mitigate environmental pollution both on short term and long term basis. The solutions developed in this research will also be extendable for environmental and soil monitoring, water quality assessment, indoor air quality monitoring for industrial applications, and video surveillance. The impact of the proposed work to academia is the development of new courses in the area of WSNs, WSVNs, and information security, integration of research outcomes into the curriculum, and workforce training for environment monitoring and analysis.