Skip to main content

Wireless Video Sensor Networks for Security Applications

Researcher: Radu Marculescu


Wireless Video Sensor Networks for Security Applications1

Nowadays, wireless sensor networks (WSN) primarily measure scalar data like temperature and pressure, monitor wildlife, or detect natural disasters such as forest fires. The sensor nodes are battery-powered, and therefore their operational capability is fundamentally limited by the energy available at the network nodes.

As technology advances, it is envisioned that large-scale, video-enabled wireless surveillance networks, that can be deployed quickly and provide accurate, real-time visual data from the field, will be a promising application in the near future.

The wireless video network (WVN) consists of a set of sensing nodes, called video nodes which are equipped with video cameras and transceivers. The use of these high-rate wireless video networks could include on-land and at-sea surveillance, video-assisted navigation, video-assisted ship management, and remote monitoring of training exercises. As opposed to WSN, the WVN bring network lifetime, throughput, delay and a combination thereof at the forefront of performance optimization. This is because WSN have more than enough capabilities to transmit data (e.g., temperature, pressure, etc.) at low rates, while the high-rate video traffic in WVN makes throughput and delay as important as the network lifetime. These issues are described next.

Performance Tradeoffs in Wireless Video Networks

Over the past few years, energy-aware routing has received significant attention in designing WSN2. Most algorithms can be viewed as different combinations of two basic routing approaches: Minimum Energy routing, which selects the route with the least total link energy cost, and Max-min routing, which selects the route with maximum bottleneck residual node energy. On the other hand, the WVN depend critically on supporting high-rate video traffic. Several researchers have recently attempted to establish limits for throughput and delay (as well as optimal scheduling) for such networks3. However, there are a few fundamental issues that still need to be considered:

  • Throughput-lifetime tradeoff which describes the fundamental capacity of transmitting data at high rates among nodes in WVN, while maintaining the system lifetime in the presence of limited energy per node
  • Throughput-delay-lifetime tradeoff which is important for applications with strict quality-of-service requirements.

Besides energy, delay and throughput, mobility is another essential characteristic of WVN. In presence of mobility, the additional challenges for large scale network design include:

  • Mobility, topology and traffic management
  • Simplification of network dynamics

Another possible direction to achieve a certain performance objective in WVN is to use pricing as the mechanism of controlling a network and model the performance objectives with some revenue- or costfunctions. The network can use the current price of a resource as a feedback signal to coerce the users into modifying their actions (e.g. changing the rate or route). However, the cost of retrieving the control information in a large network increases exponentially as the size of network increases. It is important to provide a comparable performance by using a pricing scheme based on average conditions (hence, slowly changing), which drastically reduces the cost of retrieving the control information. We plan to investigate such issues.

Security and Privacy in Wireless Video Networks

Over the last few years, security and privacy became a central concern in ad hoc wireless network. However, most of the research work has so far focused on providing security for routing and data content, while much less has been done for providing privacy and anonymity over these networks. This is the main focus of this second part of our proposed research.

Security and privacy are complex issues in ad hoc wireless networks. A malicious node (or attacker) can easily eavesdrop the wireless communication channels, and then infer and/or interfere with the communication. In addition, because of the mobility of the nodes and the absence of infrastructure, source and destination nodes rely on intermediate nodes to relay their data. This openness makes the nodes more susceptible to attacks.

Nodes in sensor networks are severely constrained in terms computation, storage and energy resources so using asymmetric cryptography (e.g. RSA algorithm and Diffie-Hellman key agreement) is often too expensive. A promising approach is to use more efficient symmetric cryptographic alternatives (e.g. AES cipher or HMAC-SHA-1 message authentication code) which, in contrast to asymmetric cryptography, are three to four orders of magnitude faster to compute. While efficient secure schemes of data encryption, authentication and key distribution have been proposed for resource-constrained wireless environments4, these issues will become even more complicated in WVN where the resource-constrained nodes are unlikely to afford, at the same time, their primary functions (i.e. video capturing, video processing and high-rate wireless communications) and intense computations for encryption/decryption.

To deal with such issues, our plan is to focus on the privacy issue and to design an efficient anonymous routing scheme (i.e. transparent routing), which prevent attackers and intermediate nodes from knowing who transmits a packet to whom. Our approach targets basically the network layer, but it will likely consider the data-link layer for energy efficiency purposes. More than this, our approach will be different from previous work5 and provide a solution with very small overhead. In addition to design the transparent routing algorithm, we also plan to propose a way to analyze its performance and compare it with other information-privacy schemes. The ultimate goal is to determine the fundamental limit of throughput-lifetime tradeoff, while considering information security and privacy.

Anticipated impact

Video sensor networks represent an emerging design platform with countless applications and profound impact on our life. The design methodology we plan to develop can be utilized to explore the design space for efficient implementation and guide the design process based on precise design metrics and cost functions. The resulting implementations will be faster, lower power, more.