The future social metaverse traffic demands over 5G and beyond (B5G) networks can be supported by efficiently placing the user plane function (UPF). Third-generation partnership project (3GPP) standard mandates that all application traffic in B5G core network must be routed through a UPF before it can reach its destination. In this regard, we propose an optimal UPF placement algorithm, referred to as Decomposed Minimum Power UPF Placement (DeMPUP), which selects the minimum power consumption path from the next-generation node B (gNB) to the edge server on the overlay backbone network. Coupled with the optimal path selection, the proposed algorithm also identifies the optimal node over which the UPF would be placed. We provide a power consumption model of the core network with a corresponding joint non-linear programming (NLP) minimization problem for UPF placement. Owing to the high complexity of the formulated NLP, for enhanced scalability over large B5G cores, we decompose the problem into two sub-problems. The first sub-problem identifies a set of feasible paths commensurate with UPF power requirement, and the subsequent sub-problem selects the optimum path among the feasible paths that minimizes the power consumption. The selected path also uniquely identifies the node to place the UPF for minimum power consumption. Analysis of DeMPUP shows that it executes in linear time and scales linearly with the number of nodes. Experimental results reveal significant power reduction for different network sizes.

This work proposes routeNow, a routing algorithm designed for wireless Software Defined Internet of Things by dynamic path relaxation based on real-time heuristic that always gives the best least hop path. This is aimed towards providing an alternative backup network for a smart hospital that functions independently on the failure of the primary hospital network. Given a pair of source and destination nodes, a central controller with a global network view first calculates a minimum hop path greedily by minimizing the remaining distance to destination. Hence, the path so found is relaxed dynamically by substituting certain links with other links keeping the hop count constant if a new path is found with a lesser latency than the previous one. For this, the controller maintains a matrix containing real-time delay between neighboring IoT nodes given the current network state and whose values are periodically refreshed using a heuristic function based on actual delay measurements. This enables the controller to dynamically modify routing paths. Simulation analysis of routeNow shows that it formulates routing paths with significant less time even with large networks and by always maintaining a route with least hops, on average a least latency path is also ensured.

We propose xDIoT, a paradigm for reliable cross-domain communication across separated IoT network domains over some public network such as the Internet. Depending on use-case scenarios, several individual IoT domains, each consisting of heterogeneous end-devices (sensors and actuators) are deployed across geographical regions. When these domains need to communicate with one another they can use an intermediate public network like the Internet. To this end, a standard paradigm is required for such inter-domain communication over public networks to reduce latencies and prevent inconsistencies, which is absent in current deployments. With xDIoT we address this issue to provide a uniform communication paradigm. In xDIoT, each individual domain has an associated gateway access point (AP) acting as the bridge between the intra-domain IoT network and the external public network. These APs perform Domain Information Exchange through JSON format to gain knowledge about each other and use a generalized packet header structure to encapsulate all data flowing between these APs. Through the use of JSON data exchange and the proposed packet header, the APs can perform seamless inter-domain communication. Implementation and analysis show that xDIoT achieves about 10% improvement in total communication latency with 80% improvement in packet processing time at individual gateway APs.

We propose an idea to speed up instruction execution through a probabilistic approach, using the parallelism offered by quantum computers. For this, we divide the instruction set of an arbitrary quantum instruction set architecture (QISA) into separate groups and then bias certain qubits representing the group so that only the instructions within the group have a high probability of getting executed in a quantum processor. Therefore, the result generated will be the superimposition of the qubits as if all the instructions within the group were executed simultaneously. We show that we can achieve a significant design improvement compared to classical computer.