Lyapunov Optimization Framework for 5G Mobile Nodes With Multi-Homing

5G Terminals with Multi-Streaming Features for Real-Time Mobile Broadband Applications

“…1 [30]. However, despite the presented system model and AQUAplus optimization algorithm given in [30], the novelty of our framework presented in this paper is the development of novel adaptive QoS Module with advanced QoS user-centric network aggregation algorithm using vertical multi-homing and multi-streaming features for delay sensitive, real-time broadband services (applications).…”

“…1 [30]. However, despite the presented system model and AQUAplus optimization algorithm given in [30], the novelty of our framework presented in this paper is the development of novel adaptive QoS Module with advanced QoS user-centric network aggregation algorithm using vertical multi-homing and multi-streaming features for delay sensitive, real-time broadband services (applications). We refer to it as Massive throughput Advanced QoS-based User-centric Aggregation (MassiveAQUA) algorithm, which is defined independently from the wireless and mobile technologies and is doing Lyapunov driftplus-penalty optimization for packet optimal scheduling in the 5GT with multiple RAT interfaces (in order to achieve massive throughput), achieving queue stability and simultaneously minimizing the queue delays, consequently having minimal overall end-to-end delay.…”

“…The 5GT system model ( Fig. 1) is almost the same as the system model in [30], with several differences: different utility functions (enriched with the delay constraints and penalties), different Lyapunov function, and existence of a packet drop decision vector for the network of queuesleading to delay sensitive and optimal multi-streaming and multi-homing processes. All novelties and differences with the AQUAplus algorithm optimization process are given in the following sections.…”

“…Each source arrival process x i (t) after passing over the transport layer, is entering in the dedicated Network Layer Processor NLP i inside the Processors part of the Upper Network layer. Despite the real M queue network in [30], here we are considering M queue network with queue vector Q(t) =(Q 1 (t), Q 2 (t), …, Q M (t)) that evolves in slotted time t  {0,1,2,3,…} with the following update equation: each queue we are trying to achieve maximal output serving rate μ j (t) on each RAT interface j = 1,2,.., M, so the sum of all output serving rates, over the time, it will be also with maximal value. For the network of queues we are introducing a packet drop decision vector …”

Cross-layer control for bounded shared state inconsistency in wireless IoT devices

“…Emerging 5th Generation wireless systems (5G) concepts are embracing these challenges faced by IoT devices [3]. Proposed 5G standards consider device limitations and intermittency, multiple medium access methods [4], Device-to-Device (D2D) communication, and new radio-access technologies [5]. The propositions of 5G are more about IoT and enabling mission critical applications than higher throughput.…”