Article Preview
Top1. Introduction
The Internet of Things (IoT) is a network of devices with unique Internet protocol addresses and software components forming an efficient distributed workflow (Olakanmi & Odeyemi, 2020b). Meanwhile, the two-layer architectural framework of IoT affects the scalability and security of the IoT networks (Olakanmi & Odeyemi, 2020a; Olakanmi & Odeyemi, 2020b; Yang et al., 2018). Most times, IoT devices are directly connected to the network or behind outdated operating systems or firewalls, exposing the network to malicious nodes. With these, an adversary can launch industrial espionage or destructive attacks on a critical infrastructure. Asides from this, the profligacy of attack on privacy couple with the increasing awareness of users to the implications of privacy compromise is affecting the seamless adoption of IoT (Castiglione et al., 2016). The large volume of traffic on the band limited communication networks, due to the exponential increase in the number of connected IoT devices, is also affecting the performance of the IoT networks. Although the advent of 5G and 6G of mobile wireless networks technologies increases the network bandwidth and speed, the apathy towards the development and adoption of these technologies is inhibiting their efficiency (Jesse, 2019; Will 5G Be Bad for Our Health, 2019). With the predicted 26 billion IoT devices in 2030 and the public apathy towards 5G and 6G technologies, it has, therefore, become imperative to develop an approach to solve the IoT networks’ security and traffic issues (Olakanmi & Odeyemi, 2020b).
Several solutions have been proposed to combat these issues (Chien-Ming et.al 2012; Lin et. al 2013; Olakanmi, 2017; Zhong et al,. 2018; Lu et. al, 2017; Rahman et al., 2016; Lu et. al., 2017; Tan et al., 2018; Olakanmi & Odeyemi, 2020b; Li et al., 2010; Lei & Jing, 2017; Li et. al, 2012; Olakanmi & Adedamola, 2019a, 2019b). However, the peculiarities of the IoT networks such as low computational power of the IoT devices, band limitation of their communication channels, intermittent generation of voluminous data by IoT devices, and energy constraint of the IoT devices made the adoption of some existing solutions infeasible for IoT networks. For example, some existing security solutions adopt bilinear pairing and homomorphic encryption which increase their complexity in terms of computational power and energy requirement. Also, a few of the schemes do not have the perfect recoverability feature for an effective de-aggregation (Olakanmi & Odeyemi, 2020b).
To solve these challenges, we come up with an improvement on our previous work in (Olakanmi & Odeyemi, 2020b; Olakanmi, 2017). Our approach groups the sub-network’s data into two; heterogeneous and homogenous data. For the homogenous data, the cluster head (CH) computes the mean of the homogeneous data as the session data for all the devices in the sub-network, while the heterogeneous CH aggregates all the heterogeneous data using the proposed recoverable aggregation approach. Highlighted below are the contributions of this work.
- 1.
Smart data recoverability approach is developed for the aggregation of the heterogeneous data to reduce the traffic on the communication network. Recoverability is one of the elusive performance characteristics of a data aggregation scheme for heterogeneous devices.
- 2.
Filtering out malicious nodes and their activities is vital to the performance of the Internet of Things. To achieve this, we proposed a non-complex batch signature verification approach for the heterogeneous data. The signature is capable of authenticating data, classifying different sub-networks, and isolating injected data.
- 3.
A secure in-and-out-data sharing feature is developed to prevent unauthorized data access.