The Internet of Things (IoT) has been used in a variety of fields, but its full potential has yet to be realized due to challenges such as scalability, security, privacy, connectivity, and so on. The commerce industry is another area where IoT devices are widely used. When a product or service is purchased (e.g., vehicle charging, parking payment, vending machine purchase, etc), IoT devices will need to send and receive payments at some point. Crypto currency-based payments [1] [2] provide a higher level of privacy and security for both parties by concealing the payee and payer identities, preventing fraud through the use of cryptographic techniques, and providing non-repudiation in the event of a dispute. A consensus method (e.g. Proof of Work, Proof of Stake, etc.) is used in Blockchain technology [3], which eliminates the need for a central authority to authorize and maintain records. The failure of a participant does not cause the system to collapse because the Blockchain ledger is irreversible and anybody can have a copy of it. It not only eliminates the problem of a single point of failure, but it also allows for secure transactions in an untrustworthy environment. The distributed structure of the Blockchain is the source of the strength it holds, which unfortunately becomes the point of weakness in scalability [4], [5] when it comes to increased number of users and payments. Bitcoin's design, in particular, renders it inherently time-consuming and slow. For example, adding a new block to the Bitcoin network takes about 10 minutes by design. Furthermore, the block size limit has an impact on performance. The number of transactions that may be completed in a given time frame is limited not only by these design characteristics, but also by hardware and bandwidth constraints. The theoretical maximum number of transactions per second is calculated to be 7 represents in the [6], which is far less than what Visa and MasterCard can process [7]. Furthermore, the transaction charge, which can skyrocket on busy days [8], is disproportionately high in comparison to the amount being sent. As a result, hefty transaction fees and long block confirmation times are two major roadblocks to virtual currencies growing and becoming widely utilised for micropayments in everyday life. One of the solutions proposed to alleviate the difficulties of virtual currencies is a payment channel network (also known as off-chain networks) [9]–[11]. It makes use of the smartcontract concept to avoid having to write each transaction to the blockchain. Rather, transactions are carried out off-chain. Basically, once a channel is created between two parties, an infinite number of transactions can be performed in both directions as long as there are available funds. Furthermore, because building a channel is a time and money-intensive procedure, nodes in a payment channel network can send payments to any other node by paying a transaction fee to other existing nodes. This creates an overlay network to simulate the payment channel network, with the nodes' balances acting as links. The Lightning Network (LN) is a recent illustration of this notion, having grown to approximately 10,000 subscribers in less than two years [9]. This payment network has nodes which charge transaction fees to users passing their data over them. The PCN concept will open up new possibilities for users and enterprises in the IoT sphere. An automobile (light node) can, for example, make a payment by connecting to the LN through a full node in the hypothetical design shown in Fig. 1. Because the entire LN node maintains the LN protocol, this concept, which serves as the foundation for the startup company Breez [9], may be adapted to any use case. While establishing a payment channel network is a promising option for addressing the scalability issue of blockchain-based cryptocurrencies for micropayments, it comes with its own set of hurdles in terms of operational efficiency, management, and routing, among other things.
One of the most distinguishing features of these payment channel networks is the manner in which channel capacities are spent, which poses new issues in terms of channel balances during routing. For example, the most well-known application of this notion is LN [9], in which nodes are free to set transaction fees, which are used in conjunction with channel capacity to select sender-to-receiver paths. Obviously, consumers choose routes that have available capacity and charge the least amount of money. As a result of this collection, available funds in one direction may be depleted, resulting in a graph that is poorly connected or even unconnected (i.e., partitioned). As a result of unidirectional payments and a lack of awareness of imbalanced channels, non-conductive nodes emerge, reducing the overall efficiency of payment routing. For example, according to one of the most recent studies, the chance of sending a $5 payment in LN [9] is around 50%, which is unacceptable to consumers. Unfortunately, this issue has received little attention, as the focus has primarily been on routing techniques. In the long run, we believe that balance-aware routing can increase the overall network's stability and success ratio. In the long run, we believe that balance-aware routing can increase the overall network's stability and success ratio. We suggest the use of two unique strategies in this study to overcome the concerns with unbalanced routing in payment channel networks. First, rather than allowing users to choose their own channel weight policy; we adopt a common channel weight policy that all nodes can adapt. The term "weight" refers to a statistic based on channel balance that is employed in Dijkstra's algorithm to discover the cheapest way. Fundamentally, the goal is to encourage nodes to use high-balanced channels for payments and avoid low-balanced ones. This will make it easier to use the channels in a way that maintains available balances in all directions. Second, in order to optimise the solution further, we recommend utilising various ingress points to the network from customers (i.e., IoT devices). In this approach, an IoT will be able to select the node from which to initiate a payment. Multiple access points will be beneficial, especially if the payment flow is skewed. These two elements will result in symmetrically balanced channels and increased payment channel network efficiency. We implemented and tested the proposed approach's effectiveness in a variety of payment circumstances, and found that the payment routes can be greatly balanced. The following is a breakdown of the paper's structure: Sect. 2 highlights previous research in the field, while Sect. 3 gives background information on payment network principles and our assumptions. The problem specification and our methodology are presented in Sect. 4. In Sect. 5, evaluate the suggested mechanism's performance. Finally, Sect. 6, concludes the entire research work.