Reliability, Scalability and Security in Smart Utility Networks
Type of Degreedissertation
MetadataShow full item record
With the rapid increase in deployment of smart meters across North America, resource consumption data of utilities such as water, gas and electricity is being collected at a much higher granular level. The huge amount of data is being used to make demand-response applications smarter. A large part of this deployment of smart utility networks and Advanced Metering Infrastructure (AMI) is being done using wireless mesh architecture due to the low deployment costs offered by this method. Although wireless environment parameters such as fading and path loss differ widely from home, outdoor to industrial, in-building scenarios, efficient protocols at the Medium Access Control (MAC) and Physical (PHY) layers and lower layer agnostic data routing protocols can be employed to overcome these challenges. For such routing protocols to be applicable to smart utility networks, reliability, scalability and security are vital metrics which will determine their performance in such networks. Our objective in this dissertation is to present a Reliability, Scalability and Security (RSSSUN) suite of recommended protocols most applicable for Smart Utility Networks. We present the performance analysis of wireless mesh routing protocols most applicable to smart utility networks. New and emerging routing protocols being proposed as an alternate standard by the Internet Engineering Task Force (IETF) community have been chosen for performance comparison. Herein, we analyze the reliability of three routing protocols, viz. RPL (IPv6 Routing Protocol for Low power and Lossy Networks), LOAD (6LoWPAN Ad Hoc On-Demand Distance Vector Routing) and a proprietary flavor of Geographical Routing developed by Landis+Gyr. Routing metrics of packet success probability, end-end delay, hop count, link quality and packet delivery ratio have been considered for this performance analysis. Further, realistic network topologies, obtained from actual smart meter network deployments in North America have been modeled in simulations to derive at results pertinent to the applicability of these wireless mesh routing protocols to Advanced Metering Infrastructure comprising of smart utility networks.With respect to scalability, we address the issue of scalable routing protocols being vital to its successful deployment in smart utility networks. To that effect, we present two approaches to modeling the wireless mesh network with the goal of analyzing the scalability of routing protocols in such networks with respect to large scale deployment. Approach I models the network as being connected with a Poisson distribution with density and transmission range as parameters. We quantify an upper bound to the network size at a maximum total expected successful packet transmission as determined by the packet success probability of each link. We validate these results with simulation data from a large scale network using the supercomputer infrastructure at the Alabama Supercomputing Authority located in Huntsville, AL. In Approach II, we model the wireless mesh network traffic arrival process as a Markov Modulated Poisson Process (MMPP) with two distinct modes. Further, an MMPP(2)/M/1/N queuing model is analyzed with the same goal of finding a network size upper bound, such that stability is maintained in the network. Verification of the model with analytical and simulation data is presented with conclusive scalability bounds affecting performance With respect to Security, we present the urgent need to implement state-of-the-art cryptographic schemes to devices in smart utility networks. Further, we entail the security risks and possible mitigation practices. Two popular industry adopted public-key cryptographic schemes of RSA (Rivest-Shamir-Adleman) and Elliptic Curve Cryptography (ECC) are compared for performance on the Tmote-Sky hardware platform. We present the impact and trade-offs involved in implementing security, an indispensable requirement in smart utility networks.