Increasing spectral efficiency has been a focal point in cellular network studies for a long time. However, a new metric, energy efficiency has recently been introduced in order to increase spectral efficiency and decrease energy consumption. In this dissertation, we investigate energy-efficient mechanisms in Heterogeneous networks (HetNets), including scaling laws for varying cell radii, small cell deployment, adaptive range expansion bias control in HetNets, and range expansion bias control in HetNets based on C-RAN architecture.
We first investigate the effects of cell size on energy saving and system capacity and then show the effectiveness of small-cell based future mobile communication systems in terms of EE. If the path-loss exponent increases and the cell-radius ratio becomes smaller, the per-energy capacity ratio becomes larger. If the path-loss exponent is set to 4 and the per-energy capacity of a macro-cell is normalized to 1, the per-energy capacities of a micro-, pico-, and femto-cell are 16, 10^4, and 10^8 in downlink and 64, 10^6, and 10^12 in uplink, respectively. The amount of CO2 emission of one BS transmitter is approximately 181[kg] in a year, if the cell radius is set to 1[km]. If the cell radius is 500[m], 100[m], and 10[m], the amount of CO2 emission in a year is reduced to 45.25[kg], 1.81[kg], and 18.1[g], respectively.
We also investigate the asymptotic performance of energy related metrics such as BS energy consumption, capacity per user, and EE per user under several simplified assumptions and characterize their scaling laws for varying cell radii in downlink cellular networks. The scaling laws of the BS transmit power per user, capacity per user, and EE per user follow R^(γ-2), R^(-2), and R^(-γ), respectively, where R is the cell radius and γ denotes the path loss exponent.
However, a small cell has a disadvantage in its coverage. Since a small cell has a limited coverage area, a large number of small cells are needed to cover a large area and it results in an increase in the deployment cost. Furthermore, high mobility users in small cells may suffer from frequent handovers. As an alternative approach, a heterogeneous network(HetNet) consisting of large cells and small cells is introduced to achieve high data rates, high EE, and high cost efficiency through this topology. We propose an energy- & cost-efficient deployment strategy in a dense user density, high required data rate, and close-set small cells environment in HetNet topology. The previous related studies assumed that small cells are deployed only in the large cell edge areas to support low-SINR users in that area. Actually, in the future mobile communication environment, we need more small-cells as pico-cells and femto-cells. We propose a spectral-, energy-, and cost-efficient small cell deployment scheme for maximizing the following spectrum, energy, and cost related metrics: spectral efficiency, spectral and energy efficiency, and deployment cost efficiency. We find the optimal values of these metrics for varying the number of small cells and the transmit power of small cells.
We investigate small cell operation schemes based on inter-cell interference coordination (ICIC) schemes among multiple small cells and a large cell in HetNet topology. Range expansion (RE) techniques have been proposed as an ICIC scheme in HetNEts operating with the same frequency band. We propose two RE bias control schemes: a common RE bias control scheme and a cell-specific RE bias control scheme. We investigate a heterogeneous cellular network in terms of throughput and EE in order to analyze the proposed schemes. To be specific, we first derive a closed-form expression of the user throughput and the EE based on stochastic geometry with a Poisson point process (PPP) theory. Furthermore, we analyze the HetNet with an adaptive RE scheme for varying active user densities under the consideration of sleep-mode operation of small BSs and find the optimal RE bias based on the analytical model. Numerical results show that the proposed bias control schemes yield 70% higher performance improvement in terms of both user throughput and EE, compared with the conventional adaptive RE control schemes.
Finally, we investigate RE bias control of small cells in HetNet topology based on a C-RAN architecture. We propose an RE control scheme of remote radio heads (RRHs) because the RE is the most commonly used scheme in HetNets. First, we compare the conventional non-C-RAN structure and a C-RAN structure. The performance of the C-RAN structure shows approximately 123% higher network throughput and 141% higher EE, compared with the performance of the non C-RAN structure at an active user density value of 50[km^2]. We also investigate the characteristics of the C-RAN structure. If we use the C-RAN structure, we can use more small cells or RRHs to improve system performance such as EE because the interference is mitigated among cooperating RRHs. Furthermore, the C-RAN structure yields a higher performance gain through the proposed RE bias optimization scheme than the non-C-RAN structure. This effect is caused by CoMP scheme among RRHs. Finally, we find the optimal value of the RE bias for varying the active user density. Through the proposed optimization scheme, we can achieve approximately 20% higher network EE, compared to the non-optimization scheme at an active user density value of 50[km^2].
스펙트럼 효율을 증가시키는 일은 오랫동안 셀룰러 네트워크 연구의 초점이 되고 있다. 그러나 최근, 새로운 메트릭인 에너지 효율이 스펙트럼 효율을 증가뿐만 아니라 에너지 소모 감소룰 위해 도입되었다. 본 학위논문에서는 셀 사이즈에 따른 에너지 효율 분석, 소형셀 설치 기법, 이종망 환경에서의 적응적인 영역 확장 바이어스 조절 기법, C-RAN 기반의 이종망 환경에서의 영역 확장 바이어스 조절 기법 등, 이종망 환경에서의 에너지 효율적 인 기법들에 대한 전반적인 내용을 다루고 있다.
먼저 소형셀과 대형셀에 따른 파워 소모, 용량, 에너지 효율에 대한 비교를 전반적으로 다루고 이후 셀 사이즈에 따른 스케일링 법칙을 셀 반경 및 경로손실지수의 함수로 구한다. 이를 통해 우리는 소형셀이 대형셀이 비해 용량이나 에너지 효율적인 측면에서 훨씬 더 좋은 성능을 얻을 수 있다는 것을 알 수 있다.
하지만 소형셀로 전 영역을 커버하기 위해서는 너무 많은 소형셀이 필요하게 되고 이는 셀을 설치하는데 있어 더 많은 비용을 초래하게 되며, 높은 속도를 갖는 사용자가 존재할 경우에는 소형셀 환경에서 너무 많은 핸드오버를 겪게 된다는 단점이 있다. 따라서 이를 해결해 주기 위해 대형셀과 소형셀이 중첩되어 존재하는 이종망 구조가 제안되었다. 우리는 먼저 이종망 구조에서의 효율적인 소형셀 배치 기법을 연구한다. 또한 이러한 배치 기법 하에서 최적의 소형셀 갯수 및 소형셀의 전송 파워를 찾는다.
또한 소형셀의 운용방식에 초점을 맞추어 영역 확장이라고 하는 기법을 적응적으로 적용할 수 있도록 영역 확장 바이어스의 적응적인 제어를 통해 이종망 시스템에서 더욱 효율적인 소형셀 운용방식을 찾는다.
더 나아가 이러한 영역 확장 바이어스의 제어를 C-RAN 구조 기반의 이종망 환경에 적용하여 더욱 에너지 효율적인 소형셀 운용방식을 찾는다.