Instead of achievable rate in the conventional association, we utilize the long-term rate to design two types of association strategies for load balancing in heterogeneous cellular networks (HCNs). In the first type, we try to balance the network loads by maximizing the weighted sum of long-term rates. Based on this type, we introduce a power control (PC) to mitigate the network interference and reduce the energy consumption in the second type. Since the long-term rate of any user associated with some BS mainly depends on the achievable rate received by this user from the selected BS and the load of this BS, it can be used as a key factor of association design for load balancing in HCNs. To solve the first association problem, we design a one-layer iterative algorithm by converting the original sum-of-ratio form into a parameterized polynomial one. By combining it with a PC algorithm, we give a two-layer iterative solution for the second problem. Specifically, the outer layer performs a user association (UA) using the algorithm mentioned in the first type, and the inner layer updates the transmit power of each BS using a power update function (PUF). At last, we give some convergence and complexity analyses for these proposed algorithms. As shown in numerical results, the proposed strategies have a superior performance than the conventional association, and the second association strategy achieves a higher load balancing gain and energy efficiency than conventional scheme and other offloading strategies.
The densification of small cells in heterogeneous networks (HetNets) causes huge energy consumption and severe network interference. To fully exploit the potential of new network architecture, the cell selection (CS) in such HetNets should couple with reducing power consumption and network interference. To this end, we jointly perform cell activation and selection (CAS) to maximize the network energy efficiency (EE) under users’ long-term rate constraints. The formulated problem is in a mixed-integer fractional form and hard to tackle. We need to transform it into a parametric subtractive form, by which we reach its solution through a three-layer iterative algorithm. The first layer searches an EE parameter using bisection method; the second layer alternately optimizes CAS indices; the third layer solves CS and CA (cell activation) problems using dual decomposition and fixed point iteration respectively. At last, we give some complexity and convergence analyses for the designed algorithm, and investigate the impacts of different network parameters on system performance. The simulation results show that the CA introduced in CS is a good option to reduce energy consumption and network interference.
With the emergence of more and more applications of Internet of Things (IoT) mobile devices (IMDs), a contradiction between mobile energy demand and limited battery capacity becomes increasingly prominent. In addition, in ultradense IoT networks, the ultradensely deployed small base stations (SBSs) will consume a large amount of energy. To reduce the network-wide energy consumption and prolong the standby time of IMDs and SBSs, under the proportional computation resource allocation and devices’ latency constraints, we jointly perform the device association, computation offloading, and resource allocation to minimize the network-wide energy consumption for ultradense multidevice and multitask IoT networks. To further balance the network loads and fully utilize the computation resources, we take account of multistep computation offloading. Considering that the finally formulated problem is in a nonlinear and mixed-integer form, we develop an improved hierarchical adaptive search (IHAS) algorithm to find its solution. Then, we give the convergence, computational complexity, and parallel implementation analyses for such an algorithm. By comparing with other algorithms, we can easily find that such an algorithm can greatly reduce the network-wide energy consumption under devices’ latency constraints.
1 October 2021 in IEEE Transactions on Vehicular Technology
To meet the demand of green communications of ultradense mobile devices (MDs), an energy-efficient mechanism, jointly considering the computation offloading and resource management, is designed to minimize the network-wide weighted energy consumption under the delay constraints of MDs for ultradense heterogeneous networks. Such a mechanism tightly integrates with the adjustment of computation capability and transmission power of MDs. Considering that the finally formulated problem is in a nonlinear and mixed-integer form, we design an effective algorithm to solve it. Specifically, by utilizing the powerful global searching capability of genetic algorithm (GA) and the accurate local searching capability of particle swarm optimization (PSO), the adaptive GA with diversity-guided is firstly used for coarse-grained search, and then adaptive PSO is utilized for fine-grained search. After that, some detailed analyses on the convergence, computation complexity and parallel implement are provided for algorithms. Simulation results show that the designed algorithm can achieve a lower network energy consumption than other offloading algorithms in general. At the same time, the numerical simulation also reveals that the designed algorithm may be more suitable for ultradense networks than existing algorithms.
In this letter, we design an association strategy to maximize the weighted sum energy efficiency (EE) for massive multiple-input and multiple-output (MIMO)-enabled heterogeneous cellular networks. Considering that the final formulated problem is in a sum-of-ratio form, we first need to transform it into a parametric nonfractional form, by which we can achieve its solution through a two-layer iterative algorithm. The outer layer searches the EE parameters and the multipliers associated with signal-interference-plus-noise ratio constraints using Newton-like method, and the inner layer optimizes the association indices using Lagrange multiplier method. Then, we give some convergence and complexity analyses for the proposed algorithm. Numerical results show that the proposed scheme significantly outperforms the existing one on the system throughput and network EE under a certain condition. In addition, we also investigate the impacts of the number of massive antennas and the transmit power of each pico base station on the association performance.
To meet the need of computation-sensitive (CS) and high-rate (HR) communications, the mobile edge computing and caching framework has been widely regarded as a promising solution. When such a framework is implemented in heterogeneous cellular networks (HCNs), it is a key and open topic how to assign mobile edge computing and caching servers to mobile devices (MDs) with CS and HR communications. In this article, under the resource partitioning and resource constraints, joint MD association and resource allocation is performed to minimize the total delay of MDs. In such a formulated problem, the MD association and resource allocation parameters are in a coupling form. In addition, it is also in nonlinear mixed-integer one and hard to tackle. To solve it, we try to develop an alternating optimization algorithm according to the coalitional game and convex optimization theorems. As for the designed algorithm, we provide some detailed convergence, stability, parallel implement, and computation complexity analyses. At last, in the simulation, we introduce other association algorithms for comparison, and investigate the impacts of MD density, time, and frequency partitioning factors on the delay of MDs. Simulation results show that the designed algorithm can achieve lower network delay than existing algorithms in general.