A QoS-Based Cross-Tier Cooperation Resource Allocation Scheme Over Ultra-Dense HetNets

Abstract

An ultra-dense network (UDN) can increase the system throughput by deploying a mass of low-power nodes and can greatly increase the spectral efficiency and energy efficiency at local hot spots. However, due to the random deployment of a large number of base stations (BSs), severe inter-cell interference may occur, which hinders the development of resource allocation technology, especially on the computational complexity. In this paper, we propose a quality of service (QoS)-based cross-tier cooperation transmission scheme over the UDN. In order to mitigate the interference and improve the entire network throughput, a non-convex problem is formulated. To solve this problem, first, we define the cooperation signal-to-interference-plus-noise ratio (SINR) threshold that represents the ratio of transmission power of macro BS and SBS, and we utilize the stochastic geometry theory to analyze the effect of cooperation SINR threshold on the number of user equipment. Then, we use the dedicated channel to decrease the cross-tier interference according to the fractional frequency refused mechanism. To reduce the computational complexity, this resource allocation scheme is divided into two parts, i.e., subcarrier allocation and power allocation, and we also consider the QoS requirement. The simulation results verify that our proposed scheme has better performance than the other two existing algorithms while QoS is guaranteed. INDEX TERMS Cross-tier cooperation, dedicated channel, quality of service (QoS), system throughput, ultra-dense network.

A QoS-based joint user association and resource allocation scheme in ultra-dense networks

“…Then, f (p) and g(p) are obviously two concave functions. Thus, utilizing the structure of objective function, the DC programming method [21][22][23][24][25] can be applied to convert the objective function of (13) into f (p) − g(p) . In the similar manner, C5 can be written as (13)…”

“…We initialize p 0 to a column vector with NL elements, where the fix power allocation of each PRB in our algorithm 2 is used as the initial power p 0 . (24) f p (τ +1) , where the complexity of graph construction algorithm is O(NKL 2 N 2 max ), and the complexity of PRB allocation algorithm is O(L 2 ) . As a comparison, the complexity of exhaustive search algorithm for the optimal solution is O(K L ) and the complexity of SA algorithm or RA algorithm is O(L) .…”

“…[20] formulates a graphbased joint user-centric overlapped clustering and resource allocation problem, in which traffic-load and the limited number of PRBs are considered. In order to improve system throughput and reduce computational complexity, [21][22][23][24][25] utilize the DC programming technology, which can be used to convert a non-convex function into the difference of two convex functions, and to obtain an approximately optimal solution through an iterative method. However, the joint user association and resource allocation scheme under the virtual cell architecture, considering the use of multiple PRBs for each user to reduce interference, cell load balance, guarantee QoS requirements and improve system throughput, has not been studied in previous work.…”

A Cluster-Based Energy-Efficient Resource Management Scheme With QoS Requirement for Ultra-Dense Networks

“…To guarantee the user experience, [20]- [30] further take the QoS requirement of UEs into account. Although QoS-aware algorithms based on convexoptimization are investigated to realize different targets, such as system capacity maximization [20], [21], EE maximization [22], the sum of transmits power for all BSs minimization [23], [24], and SE maximization [25], most of these algorithms are unsuitable for large scale UDN due to the high computational complexity. To better apply to UDN, [26] proposes an interference-separation clustering-based scheme to divide the massive small cells into smaller groups with different priorities, which reduces the network scale.…”

A Model-Driven Deep Reinforcement Learning Heuristic Algorithm for Resource Allocation in Ultra-Dense Cellular Networks

“…The convex optimization based methods normally transform a nonconvex mixedinteger problem into some approximated convex sub-optimal problems which may be solved locally with a certain number of iterations [6][7][8][9]. Although current approximation methods demand low computational complexity, they are usually assumed to know the perfect knowledge about the channel state information (CSI) [10,11], which may be impractical for a dynamic wireless environment due to the huge overhead for feedback channel. In addition, the discrete optimization assisted approaches mainly utilize game theory or graph theory to solve the joint optimization problem effectively with acceptable amount of complexity [12,13].…”