Ph.D Projects in Wireless Sensor Networks
Wireless sensor network (WSN) has significantly emerged as a key technology for potential real-time applications in a wide range of domains. It enables new emerging areas of research and advances in smart computing technology. The prime advantage of WSN is the availability of tiny, inexpensive, and easily deployable smart sensors to expanding usage of sensing in smart environment. However, a lot of research gaps must be addressed to facilitate the widespread usage of WSN technology in real-world applications.
WSN has its application not only in the environmental monitoring and military applications, but also in smart homes, health care with the integration of IoT and cloud computing makes its presence everywhere in future smart environment. Nevertheless, it has many issues due to the factors in terms of sensor heterogeneity, energy efficiency, network, environment conditions, and application requirements.
Extension to the existing protocols or modeling a new protocol in NS2 for WSN can be done in the following aspects. The protocols are developed as C++ files and the dependent files are modified in the ns-allinone package. The directory structure of the newly added files is mentioned in Makefile. Then NS is rebuilt using Linux utility tool make. Simulation is carried out by mentioning the newly developed protocol in the TCL configuration file. Implementation aspects in NS2 of the WSN protocols are as follows.
- Deployment strategies for sensor network can be implemented in TCL script where the location of the node is configured and also sensor node configurations such as different communication and sensing range, battery configuration can be fixed in TCL script.
- Due to the energy hole in WSN, if connectivity gets affected, sensors are required to increase their transmission range to establish a communication link with far away located sensor. In such cases, transmission power control techniques are applied in physical layer protocol files.
- Mannasim, the sensor simulation framework is patched to NS package for the simulation of cluster based routing protocol LEACH.
- Clustering can also be done by modifying the on demand routing protocol files in such a way that whenever a packet for a specific destination/base station is arrived at the routing layer of the node, its destination ID changes to the ID corresponding to the clusterhead of the node, thereafter clusterhead forwards the data to the destination.
- In aggregation cases, on receiving data from the cluster members, clusterhead aggregates the data and then forward it to the destination. Clustered selection and data gathering from the sensors are done periodically by using timer class.
- Recently, geographic routing protocol is used in WSN. Hence GPSR routing protocol files can be patched to NS package and its advancement such as opportunistic routing can be done by exploiting overhearing functionality and data cache maintenance at the node by modifying GPSR files.
- Various attacks such black hole, gray hole, wormhole, flooding, misrouting, modification, jellyfish, replica, data aggregation attacks, misbehavior of clusterhead and other attacks can be modeled by modifying the behavior of the attacker nodes in corresponding protocol files.
- Attack detection techniques such as applying overhearing functionality and utilizing that in detection mechanism can be integrated into protocol functionality in the respective files.
- Attack prevention security algorithms such as RSA, ECC, HMAC can be integrated for security related solutions. Since sensors are battery constraint light weight security algorithms such as Chinese Remainder Theorem can be integrated with security related functions.
- Parameters such as signal strength, energy, queue size, packet priority can be attached in the packets so that it can be accessed in protocols of other layers to develop the cross layer solutions.
- Network Performance in WSN varies depends on various network characteristics.
– Deterministic deployment
– Random Deployment
– Number of communication flows
– Data rate of each flow subject to the constraints of the number of channels and interfaces, and bandwidth of the channel
– Number of nodes in the network,
– Area of the network
– Distance between source and destination
– Varying number of clusters
– Size of cluster
– Clusterhead reelection period
– Number of rounds
– Communication range
– Sensing range
– Transmission power
– Battery energy
– Type of the antenna- omni or directional antenna
– Number of attackers
– Percentage of malicious behavior
– Pause time
– Type of mobility model (Random Way Point, Random Walk, etc)
– Static sink
– Mobile sink
– located at the network center
– located at the border of the network
- The impact can also be observed with the selection of the protocol at each layer. One example of this is with the same scenario, the choice of of signal propagation type such as Two Ray Ground, Free space, Shadowing in physical layer can make the difference in results.
- One such kind of variation is evaluating the performance for various data dissemination periods of sensors at the application layer protocol.
- Energy consumption
- Residual energy
- Network lifetime
- Percentage of alive nodes and dead nodes
- Aggregation accuracy and deviation
- Packet Delivery Ratio (PDR)
- Average End to End Delay
- Routing Overhead
- Control Overhead
- Storage overhead
- Packet Loss
- Hop count
- Percentage of coverage
- Attack detection accuracy
- Attack detection time
- False Alarm
There are numerous metrics depending on the proposed approach and the network scenarios. AWK script for these metrics can be applied to process the trace file that consists of the information such as event time, event type, node ID, packet sequence number, type and size of the packet, the layer at which the event occurs, reason for packet drop, remaining energy of the nodes, and TTL. The results from the execution of AWK script can be plotted as Xgraph in NS2 for the purpose of self or comparative protocol analysis.