Design and Optimization of Application-sensitive Heterogeneity-aware Traffic Engineering for WSNs

Student thesis: Doctoral ThesisDoctor of Philosophy


Wireless Sensor Networks (WSNs) support a wide range of both civil and military applications such as smart health monitoring, smart environmental reporting, surveillance of sensitive locations and intrusion detection. Implementation of these sensitive applications in WSNs demand sophisticated designs and advanced management systems to achieve commercial success. However, recent growth of industrial application spectrum challenges researchers to deal with of resource constraints of sensor’s limited energy and application dependency in traditional WSNs. In WSNs, performance of successful data delivery over longer network period depends upon heterogeneity-aware and application-sensitivity cooperations among deployed sensor nodes. Many distributed and centralized routing algorithms have been proposed to enhance the cooperations among sensor nodes of conventional WSNs. Although proposed solutions show reasonable improvement regarding network lifetime but still struggle to offer required robustness to deal with large scale application-sensitivity and heterogeneity-awareness. These solutions require careful inspection of residual resources during network operation and need skills to reconfigure the nodes according to their residual potential. Although distributed routing protocols extend networking functionality over larger network area but also cause drastic uneven cluster-formation. Meanwhile, central management system with full and partial centrally controlled routing algorithms show reasonable consistency over distribution of network resources and offer significant load-balancing. Recently, Software Defined Networking (SDN) provides many attractive solutions for central management of WSNs to offer reliable flexibility and adaptability by executing the separation of the control plane from the sensor nodes forwarding plane. This recent penetration of SDN-based solution encourage researchers to develop heterogeneity aware designs that can manage vendor-specific applications for WSNs.
In this dissertation, we develop heterogeneity-aware and application-sensitive routing protocols for both conventional and SDN-enabled WSNs architectures. First, we provide an overview of existing state-of-the art centralized and distrusted traffic engineering routing solutions for conventional WSNs and SDN-based architectures. Next, we propose homogeneity aware centralized clustering routing protocols to establish a foundation for central traffic engineering capability of WSNs. For backward compatibility to distributed clustering routing protocols we also propose a hybrid routing solution to enrich network scalability. Then finally we propose routing solutions which accommodate on-demand SDN-based central management system.
In chapter 3, we propose two energy-efficient path planning routing protocols for three level heterogeneous WSNs namely, Two-Hop heterogeneity-aware Centralized Energy Efficient Clustering (THCEEC) and Advanced heterogeneity-aware CEEC (ACEEC). In proposed models, BS utilizes algorithms of THCEEC and ACEEC routing protocols to identify the suitable Cluster-Heads (CHs) by considering initial energy, residual energy, regional flag, and nodes’ distance towards BS. The proposed algorithms also perform intelligent distribution of energy resources evenly across the heterogeneous WSNs at the initial stage of cluster formation. Extensive simulation results confirm the reliability and energy efficient performance enhancements of the centralized cluster formations of ACEEC and THCEEC, which acquire better network lifetime and successful data forwarding as compared to the conventional state-of-the art distributed routing protocols.
In chapter 4, we design a hybrid routing protocol to enable backward compatibility with distributed clustering routing protocols. In this part of dissertation, we propose an adaptive immune Multi-Hoping Multi Level Clustering (MHMLC) protocol that executes a Hybrid Clustering Algorithm (HCA) to perform optimal centralized selection of Cluster-Heads (CHs) within radius of centrally located Base Station (BS) and distributed CHs selection in rest of network area. HCA of MHMLC also produces optimal intermediate CHs for inter-cluster multi-hop communications that develop heterogeneity-aware economical links. This hybrid cluster-formation facilitates the sensors to function at short range transmission power level that enhances link quality and avoids packet drop. The simulation environments produce fair comparison among proposed MHMLC and existing state of the art routing protocols. Experimental results give significant evidence of better performance of the proposed model in terms of network lifetime, stability period and data delivery ratio.
In chapter 5, we propose SDN-Based Application-aware Centralized adoptive Flow Iterative Reconfiguring (SACFIR) routing protocol with centralized SDN iterative solver controller to maintain the load-balancing between flow reconfigurations and flow allocation cost. Proposed SACFIR’s routing protocol offers unique iterative path-section algorithm which initially computes suitable clustering based on residual resources at control layer and then implements application-aware threshold-based multi-hop reports transmissions at forwarding plane. Operation of SACFRR algorithm is centrally supervised by SDN controller reside at BS. This dissertation extends SACFIR to SDN-Based Application-aware Main-value Centralized adoptive Flow Iterative Reconfiguring (SAMCFIR) to establish both proactive and reactive manner reporting. Extensive experimental simulations-based results validate that our SDN-enabled proposed models adjust reconfigurations period according to traffic burden over sensor nodes which result into heterogeneity-awareness, load-balancing and application-sensitive reconfigurations of WSNs.
Date of Award2017
Original languageEnglish
Awarding Institution
  • Dalian University of Technology
SupervisorXiaopeng Hu (Supervisor)

Cite this