Trust is considered as one of significant and efficient mode of measuring the behavior and nature of communicating node in the network. Trust quantity is generally measured in two parameters such as 0 and 1 where 0 means no trust (the node is not trustworthy anymore in the network) and 1 means the node is highly trusted (recommended to be part of communication network). In comparison of other security and cryptographic algorithms and schemes, trust is defined as one of effective way of identifying the legitimacy of communicating node in the network. In this paper, the trust aggregation method will be used in order to measure the communicating behavior of a device.
System model
Figure 2 represents the system model of a network consists of \({D}_{c}\) number of communicating devices that are able to transmit and receive the information among each other. The \({D}_{1}, {D}_{2},\dots {D}_{n}\) are the number of devices that are considered and allowed to be a part of communication at a specific interval of time \(T.\)
System model of proposed mechanism.
In order to measure the trust or legitimacy of communicating device, number of metrics are considered to understand the working of trust computation using aggregate method. The nomenclature of the proposed mechanism is depicted in Table 2.
Trust aggregation
Number of responses can be achieved of a single request from parallel number of paths. In order to understand trust aggregation in a simple way, let us take an example, where \(X\) does not know \(Z\), he asks him friends \((Ys)\) about \(Z\). Different friends \((Ys)\) may have different ideas about that individual \(Z\). Individual A should integrate the various ideas where he has received from \(Ys\) about \(Z\) to infer a unique idea about \(Z\). Individual receive multiple recommendations and information from parallel sources where they can combine and finally decide the final judgement about an individual in order to simulate the behavior of that individual. The aggregation methods depend upon the trust’s models such as mean, fuzzy method and subjective logic combining approach. In this paper, we use the hybrid methods using mean and subjective logic aggregate methods to strengthen the trust computation process of a device.
Mean and subjective logic combining operator aggregation method
It is defined as the summation of various information received from parallel sources in a particular amount of time \(T\). If a device \(D\) has \(n\) various sources for an unknown device \({D}_{u}\), then device \(D\) should aggregate the values obtained from others. If each device \({D}_{i}\) reports some believed value Bi and trust value \({TV}_{i}\) for device \({D}_{x}\), then the resultant aggregation of believe for \({D}_{x}\) will be computed as:
$${B}_{x}=\frac{\sum_{i=1}^{n}Tr\times Bi}{\sum_{i=1}^{n}TVi}.$$
(1)
Two operators are used to maintain parallel and serial opinion using consensus and reduction operators. In case, if P trust opinion to \(Q\) has in context of information X is \({T}_{Q}^{P}=({m}_{Q}^{P}, {n}_{Q}^{P}, {\eta }_{Q}^{P} , {\chi }_{Q}^{P})\) and \(Q{\prime}s\) opinion about information \(X\) is \({T}_{X}^{P}=({m}_{X}^{P}, {n}_{X}^{P}, {\eta }_{X}^{P} , {\chi }_{X}^{P} )\) then to infer \(P{\prime}s\) opinion, the \(Q{\prime}s\) opinion about information \(X\) can be reduced as follows:
$${T}_{X}^{PQ}={T}_{Q}^{P}\otimes {T}_{Q}^{P}=\left\{\begin{array}{c}{m}_{X}^{P:Q}={m}_{Q}^{P} {m}_{X}^{Q}\\ {n}_{X}^{P:Q}={m}_{Q}^{P} {n}_{X}^{Q}\\ {\eta }_{X}^{P:Q}={n}_{Q}^{P}+{\eta }_{X}^{Q}+{m}_{Q}^{P}{\eta }_{X}^{Q}\\ {\chi }_{X}^{P:Q}={\eta }_{X}^{Q}\end{array}.\right.$$
(2)
The subscript and superscript represent the trusted and trusting values. The reduction opinion is used to decrease and increase the beliefs/disbeliefs and uncertainty in transitive chains.
Consensus operator
The consensus opinion represents the fair decision between two opinions \(P and Q\) for the information \(X\) with believes as \({T}_{X}^{P}=({m}_{X}^{P}, {n}_{X}^{P}, {\eta }_{X}^{P} , {\chi }_{X}^{P})\) and \({T}_{X}^{Q}=({m}_{X}^{Q}, {n}_{X}^{Q}, {\eta }_{X}^{Q} , {\chi }_{X}^{Q})\), their corresponding aggregation will be defined as:
$${T}_{X}^{P\phi Q}={T}_{Q}^{P}\theta {T}_{Q}^{P}=\left\{\begin{array}{c}{m}_{X}^{P\Phi Q}={m}_{X}^{P} {\eta }_{X}^{Q}+{m}_{X}^{P}\frac{{\eta }_{X}^{Q}}{{\eta }_{X}^{P}}+{\eta }_{X}^{P}-{\eta }_{X}^{P}{\eta }_{X}^{Q}\\ {m}_{X}^{P\Phi Q}={n}_{X}^{P} {\eta }_{X}^{Q}+{n}_{X}^{P}\frac{{m}_{X}^{Q}}{{n}_{X}^{P}}+{\eta }_{X}^{P}-{{\eta }_{X}^{P}\eta }_{X}^{Q}\\ {\eta }_{X}^{P\Phi Q}={\eta }_{X}^{P} {n}_{X}^{Q}+{n}_{X}^{P}\frac{{\eta }_{X}^{Q}}{{n}_{X}^{P}}+{\eta }_{X}^{P}-{{\eta }_{X}^{P}\eta }_{X}^{Q}\\ {\chi }_{X}^{P\Phi Q}={\chi }_{X}^{P}\end{array}\right..$$
(3)
The consensus operator effect is to reduce the uncertainty and improve the disbelief and belief.
$$\mathrm{Pr}\left({N}_{ri}\right)=\frac{{dt}_{i}}{\sum_{i=1}^{n}{dt}_{i}}.$$
(4)
The aggregated trust values computed between neighboring devices further strengthen the trust and legitimacy in the network. The devices’ having higher trust values are counted as most trusted and reliable that can be used to lead and coordinate the entire system.
The same method can be applied to determine the aggregation of trust value to a sink device obtained after getting the believes and trust value from its neighbors. If trust value from various neighbors to the destination node is computed as \({TrustDes}_{1}, {TrustDes}_{2}, \dots ..{TrustDes}_{n}\), and devices direct trust value defined as \({TV}_{1}, {TV}_{2}, \dots .{TV}_{n}\), the device trust \(D\) to the destination as the aggregation result is computed as:
$${TrustDes}_{N}=\frac{\sum_{i\epsilon Ne(N)}{TV}_{i}\times {TrustDes}_{i}}{\sum_{i\epsilon Ne(N)}{TV}_{i}}.$$
(5)
The trusted based blockchain enabled security mechanism for IoMT application.
The presented Algorithm 1 represents the computation and evaluation of communicating devices in the network based upon their behavior and trust values in the network. The trust of each device is computed using mean and subjective logic method. In addition, the computed trust devices are kept in blockchain for further surveillance. Repeat the step until the entire devices in the network and block the altered and untrusted devices from the network.
Blockchain along with trust-based methods
As the devices are efficiently computed according to their behavior by identifying their trust values. The intruders may further insert software in order to later the ideal devices in the network that will drastically affect the overall performance of the network. Though trusted mechanisms are able to identify the legitimacy of each device, however, it is further needed to keep surveillance to the communicating devices where intruders may not compromise or alter the ideal devices. The identification of device alteration method by the intruder should be identified at the initial stage of the communicating process. In order to overcome the mentioned limitation, the trust-based method is further combined with blockchain technology that is transparent and may keep surveillance in the network. The blockchain creation and addition of devices mechanism is further described in the subsequent sections.
Blockchain after aggregation schemes
The computed trust values of devices using aggregation methods are used integrated with the blockchain network in order to maintain the transparency and used to maintain the legitimacy and trust in the network. The devices having higher trust values are used to act as miners for further validating the incoming devices in the blockchain network.
Blockchain creation
The blockchain of trusted devices is created in the network where the devices’ having higher trust values acts as miners for further validation and verification. The hex string is generated as,
$$Var TransactionHex=transaction.build().toHex().$$
(6)
Addition of device in the blockchain
The devices having moderate trust values are keep on adding in the network after creating their hash values verified by the genesis device. The devices having lower trust values are keep on surveillance in the network where their reduction in trust values immediately gets removal from the blockchain network. The proposed mechanism is verified and validated against various performance metrices such as reliability, trust, delay and authentication.
