*3.2.3 Network coding*

A review on energy-efficient and low-latency routing protocols for WSNs without dominating the other design factors is presented by Bagyalakshmi et al. [24].

*Wireless Sensor Networks - Design, Deployment and Applications*

Fault tolerance [25] enhances the availability, reliability, and dependability of the system by ensuring the usage availability of the system without any disruption in the presence of faults. In WSN, fault tolerance is also a demanding issue due to

desynchronization, communication link errors, etc., which are provoked owing to hardware and software failures, environmental conditions, etc. Hence, fault management in WSN must be administered with additional care. Initial review works on fault-tolerant routing schemes are present in literature [21, 25–28]. Yu et al. [26] have explained issues in the fault management of WSN. Three phases called fault diagnosis, fault detection, and fault recovery for supervising faults have been proposed. In fault detection phase, an unexpected failure should be identified by the system. Literature [26–28] explains various fault detection techniques. In fault diagnosis phase, comprehensive description or model has been determined to distinguish various faults in WSNs [21] or fault recovery action. In the fault recovery phase, the sensor network is redesigned from failures or fault nodes to enhance the network performance. Fault recovery techniques have been dealt by literature [25].

The major techniques utilized for attaining energy saving, low latency, long

Cluster-based architecture is a foremost technique for effective energy conser-

Data aggregation is one of the significant methods applied to aggregate the raw data evolved from multiple sources. In data aggregation schemes, nodes receive the data, reduce the amount of data by employing data aggregation techniques, and then transmit the data to the BS. The average or minimum amount of received data are merely forwarded by the received node. This reduces the network traffic and hence low latency is achieved. However, the base station (sink) cannot ensure the

vation. In this mechanism, the network is partitioned into clusters, where the cluster head (CH) is a leader to manage the members of each cluster. Every member sensor node transmits the sensed data to their corresponding CH; then, CHs communicate the collected data to the BS. This technique avoids flooding, routing loops, and multiple routes; hence, reduced network traffic and low latency are attained. The major advantage of cluster-based architecture is that it needs less transmission power because of small communication ranges within the cluster. The CH uses the fusion mechanism to minimize the size of the transmission data. CH selection is performed in a rotation basis to balance the energy consumption in the network and improve the network lifetime. However, in cluster-based routing protocols, cluster head selection plays a critical role. Further, clustering algorithms do not consider the location of the base station, which creates a hot spot problem in multi-hop

the sensor nodes more vulnerable to failure because of energy depletions,

**3.2 Major techniques used for data collection design issues**

lifetime, and fault tolerance in WSNs are discussed in this section.

*3.1.3 Fault tolerance*

*3.2.1 Cluster architecture*

wireless sensor networks.

*3.2.2 Data aggregation*

**88**

Network coding is the same as the aggregation technique. In this technique, the nodes collect the data from neighbor nodes and combine them together by applying mathematical operations; then it transmits data to the BS. This technique improves the network throughput, reliability, energy efficiency, and scalability; it is also resilient to attacks and eavesdropping. Network traffic in broadcast scenarios can be reduced by combining several packets as a single packet rather than sending separate packets.
