[Paper Review] Joint User Association and Power Control for Load Balancing in Downlink Heterogeneous Cellular Networks

Tianqing Zhou|arXiv (Cornell University)|Jul 4, 2016

Advanced MIMO Systems Optimization25 references3 citations

TL;DR

This paper proposes two joint user association and power control schemes for load balancing in downlink heterogeneous cellular networks (HCNs), using effective rate maximization to balance traffic and reduce interference. The two-layer iterative algorithm achieves higher load balancing gain and energy efficiency than conventional schemes, with the power control-enhanced scheme outperforming others in both throughput and energy efficiency.

ABSTRACT

Instead of achievable rate in the conventional association, we utilize the effective rate to design two association schemes for load balancing in heterogeneous cellular networks (HCNs), which are both formulated as such problems with maximizing the sum of effective rates. In these two schemes, the one just considers user association, but the other introduces power control to mitigate interference and reduce energy consumption while performing user association. Since the effective rate is closely related to the load of some BS and the achievable rate of some user, it can be used as a key factor of association schemes for load balancing in HCNs. To solve the association problem without power control, we design a one-layer iterative algorithm, which converts the sum-of-ratio form of original optimization problem into a parameterized polynomial form. By combining this algorithm with power control algorithm, we propose a two-layer iterative algorithm for the association problem with power control. Specially, the outer layer performs user association using the algorithm of problem without power control, 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 the proposed algorithms. As shown in simulation results, the proposed schemes have superior performance than the conventional association, and the scheme with joint user association and power control achieves a higher load balancing gain and energy efficiency than conventional scheme and other offloading scheme.

Motivation & Objective

• Address the imbalance in user load distribution caused by conventional signal strength-based user association in heterogeneous cellular networks (HCNs).
• Improve system resource utilization and user experience by enabling intelligent offloading of users from overloaded to underloaded base stations (BSs).
• Simultaneously mitigate inter-tier interference and reduce energy consumption through joint power control and user association.
• Design a scalable and convergent algorithmic framework for practical deployment in real-world HCNs.

Proposed method

• Formulate user association as a sum-of-effective-rate maximization problem, replacing traditional achievable rate with effective rate to reflect load and quality-of-service trade-offs.
• Propose a one-layer iterative algorithm for user association without power control, transforming the non-convex sum-of-ratios problem into a parameterized polynomial form.
• Develop a two-layer iterative algorithm for joint user association and power control: the outer layer performs user association using the one-layer algorithm, and the inner layer updates BS transmit power via a power update function (PUF).
• Integrate the power update function (PUF) into the inner loop to dynamically adjust transmit power and suppress interference.
• Apply convergence and complexity analysis to validate the stability and practicality of the proposed two-layer algorithm.
• Use effective rate as a performance metric that inherently captures both user quality and BS load, enabling load-balanced association decisions.

Experimental results

Research questions

• RQ1 How can effective rate be used as a metric to enable load-balanced user association in HCNs?
• RQ2 What is the performance gain of joint user association and power control over conventional association schemes in terms of load balancing and energy efficiency?
• RQ3 Can a two-layer iterative algorithm converge efficiently while balancing system throughput and power consumption?
• RQ4 How does power control improve interference management and energy efficiency in downlink HCNs with heterogeneous BSs?
• RQ5 What is the impact of user association on the average rate and energy efficiency of edge users in overloaded macrocells?

Key findings

• The proposed JUAPCMSER scheme with joint user association and power control achieves the highest average energy efficiency among all schemes, significantly outperforming conventional association and other offloading methods.
• The scheme JUAPCMSER achieves a higher load balancing gain and energy efficiency than the UAMSER scheme (which only performs user association), due to interference mitigation via power control.
• The simulation results show that the scheme MARA (maximal achievable rate association) has the highest average rate despite the worst load imbalance, due to a few high-rate users being associated with underloaded BSs.
• The proposed two-layer iterative algorithm converges rapidly, as shown in Fig. 8, with all components (outer loop, user association, and power control) demonstrating high convergence rates.
• The average rate and energy efficiency of JUAPCMSER are superior to AUF and UAMSER, with the latter achieving higher throughput and better energy efficiency than AUF due to better load distribution.
• The performance gap between schemes widens under high load, indicating that intelligent offloading significantly improves edge user experience in overloaded networks.