An Analytical Energy Performance Evaluation Methodology for 5G Base Stations

Peesapati, S. Krishna Gowtam; Olsson, Magnus; Masoudi, Meysam; Andersson, Sören; Çavdar, Çiçek

doi:10.1109/wimob52687.2021.9606296

Cited by 14 publications

(10 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The virtual network together with the RDM prediction module construct the DT, details of which will be discussed later in Section III-C and III-D. These updated parameters and the performance metrics are used to calculate the RDM in the network using Equation (16). The calculated RDM in the DT, i.e., RDM dt , is the expected or predicted value of risk in the next time window, which is defined as the duration in which the risk is evaluated.…”

Section: A Risk-aware Sleep Mode Managementmentioning

confidence: 99%

“…The problem of deciding on the sleeping as well as the level of SMs are complex. Machine learning (ML) techniques have shown promising capabilities to efficiently handle the complexity and tune the parameters in systems with ASMs [16]- [23]. The authors in [16] evaluate the energy consumption of 5G and beyond BS with multi level SMs.…”

mentioning

confidence: 99%

“…Machine learning (ML) techniques have shown promising capabilities to efficiently handle the complexity and tune the parameters in systems with ASMs [16]- [23]. The authors in [16] evaluate the energy consumption of 5G and beyond BS with multi level SMs. The authors in [17]- [19] investigate 4-level SMs with pre-defined order of (de)activation of SMs.…”

mentioning

confidence: 99%

See 2 more Smart Citations

Digital Twin Assisted Risk-Aware Sleep Mode Management Using Deep Q-Networks

Masoudi

Soroush²,

Zander

et al. 2023

IEEE Trans. Veh. Technol.

Self Cite

View full text Add to dashboard Cite

Base stations (BSs) are the most energy-consuming segment of mobile networks. To reduce BS energy consumption, different components of BSs can sleep when BS is not active. According to the activation/deactivation time of the BS components, multiple sleep modes (SMs) are defined in the literature. In this study, we model the problem of BS energy saving utilizing multiple sleep modes as a sequential Markov decision process (MDP) and propose an online traffic-aware deep reinforcement learning approach to maximize the long-term energy saving. However, there is a risk that BS is not sleeping at the right time and incurs large delays to the users. To tackle this issue, we propose to use a digital twin model to encapsulate the dynamics underlying the investigated system and estimate the risk of decision-making (RDM) in advance. We define a novel metric to quantify RDM and predict the performance degradation. The RDM calculated by DT is compared with a tolerable threshold set by the mobile operator. Based on this comparison, BS can decide to deactivate the SMs, re-train when needed to avoid taking high risks, or activate the SMs to benefit from energy savings. For deep reinforcement learning, we use long-short term memory (LSTM), to take into account the long and short-term dependencies in input traffic, and approximate the Q-function. We train the LSTM network using the experience replay method over a real traffic data set obtained from an operator's BS in Stockholm. The data set contains data rate information with very coarse-grained time granularity. Thus, we propose a scheme to generate a new data set using the real network data set which 1) has finer-grained time granularity and 2) considers the bursty behavior of traffic data. Simulation results show that using proposed methods, considerable energy saving is obtained, compared to the baselines at cost of negligible number of delayed users. Moreover, the proposed digital twin model can predict the performance of the DQN proactively in terms of RDM hence preventing the performance degradation in the network in anomalous situations.

show abstract

Section: A Risk-aware Sleep Mode Managementmentioning

confidence: 99%

mentioning

confidence: 99%

See 1 more Smart Citation

Digital Twin Assisted Risk-Aware Sleep Mode Management Using Deep Q-Networks

Masoudi

Soroush²,

Zander

et al. 2023

IEEE Trans. Veh. Technol.

Self Cite

View full text Add to dashboard Cite

show abstract

“…In [20], we extended the analytical methodology developed in [21] to propose a dynamic Q-learning-based resource adaptation algorithm to obtain higher energy savings under varying traffic loads. As an extension to [20], in [22], we mainly addressed the challenges associated with complex long-horizon problems by developing a hierarchical reinforcement learning solution wherein different optimization strategies were implemented as a hierarchy of reinforcement learning agents.…”

Section: Related Workmentioning

confidence: 99%

AI-Assisted Multi-Operator RAN Sharing for Energy-Efficient Networks

Peesapati

Olsson

Andersson

et al. 2023

Telecom

Self Cite

View full text Add to dashboard Cite

Recent times have seen a significant rise in interest from mobile operators, vendors, and research projects toward achieving more energy-efficient and sustainable networks. Not surprisingly, it comes at a time when higher traffic demand and more stringent and diverse network requirements result in diminishing benefits for operators using complex AI-driven network optimization solutions. In this paper, we propose the idea of tower companies that facilitate radio access network (RAN) infrastructure sharing between operators and evaluate the additional energy savings obtained in this process. In particular, we focus on the RAN-as-a-Service (RANaaS) implementation, wherein each operator leases and controls an independent logical RAN instance running on the shared infrastructure. We show how an AI system can assist operators in optimizing their share of resources under multiple constraints. This paper aims to provide a vision, a quantitative and qualitative analysis of the RANaaS paradigm, and its benefits in terms of energy efficiency. Through simulations, we show the possibility to achieve up to 75 percent energy savings per operator over 24 h compared to the scenario where none of the energy-saving features are activated. This is an additional 55 percent energy savings from sharing the RAN infrastructure compared to the baseline scenario where the operators use independent hardware.

show abstract

“…Compared to other common communication scenarios, energy saving in IDS is more important since power resources are limited. There is a significant amount of research devoted to minimizing the energy consumption of networks both in fixed [5][6][7] and wireless communications [8][9][10][11][12][13][14][15][16][17][18][19][20][21][22][23]. With the commercialization of 5G technology, due to the above reasons, enhancing the coverage of the mmWave with the least amount of energy has become a critical problem that needs to be solved.…”

Section: Introductionmentioning

confidence: 99%

mmWave Coverage Extension Using Reconfigurable Intelligent Surfaces in Indoor Dense Spaces

Li,

Topal,

Demir

et al. 2023

ICC 2023 - IEEE International Conference on Communications

View full text Add to dashboard Cite

Millimeter-wave (mmWave) is widely investigated for indoor communication scenarios thanks to the available rich spectrum. However, the shortened antenna size and the high frequency make mmWave extra sensitive to blockages. Indoor dense space (IDS) is a specific type of indoor environment, where the compact geometry together with a high number of blocking objects and users make it hard to fulfill the data rate required by all of the users in the mmWave network. With the capability of redirecting signals, the reconfigurable intelligent surface (RIS) has the potential to overcome the attenuation brought by the blockage. Aside from the promising improvement in data rate brought by the RIS, the power supply for RIS is also a major concern in IDS due to the cabling and the batteries. Dynamic RIS has the capability of reconfiguring its phase-shifts to offer a higher gain in data rate with the price of consuming power. In comparison, by sacrificing the reconfigurability, static RIS does not require any power, cabling, or batteries but is expected to provide lower data rates. To find the balance between the performance and cost trade-off, the concept of self-sustainable RIS in IDS is proposed. This approach involves the utilization of specific RIS elements to harvest energy, thereby providing support for the power requirements of the RIS operation, consequently reducing the reliance on traditional cabling infrastructure.In this work, we compare the coverage extension effect brought by deploying static, dynamic, and self-sustainable RISs in the aircraft cabin which is a typical example of an IDS. To capture the propagation characteristics of a RIS in IDS, we first provide guidelines for modeling the RIS in the ray tracing (RT) simulator, and then we select the best locations to deploy RISs among three candidates. For each type of RIS deployment, we propose an optimization algorithm, which jointly configures the RIS phase-shifts and the time resources to provide the maximum equal achievable data rate for all of the users. Additionally, for the self-sustainable RIS, the working mode of each RIS element is also jointly configured such that each element is used either to reflect the incoming signal or to use the signal for energy harvesting. Based on the results, the signal propagation of a single base station (BS) can be extended from 3 rows to 11 rows by deploying static or dynamic RISs. The minimal achievable data rate is 35.4 Mbps with the static RISs and 45.3 Mbps with the dynamic RISs. The results indicate that due to the limitation of selfsustainable constraints, RISs with 16 elements are hard to cover the whole 11 rows in the considered cabin. Nevertheless, with self-sustainable RIS, 10 more

show abstract

An Analytical Energy Performance Evaluation Methodology for 5G Base Stations

Cited by 14 publications

References 19 publications

Digital Twin Assisted Risk-Aware Sleep Mode Management Using Deep Q-Networks

Digital Twin Assisted Risk-Aware Sleep Mode Management Using Deep Q-Networks

AI-Assisted Multi-Operator RAN Sharing for Energy-Efficient Networks

mmWave Coverage Extension Using Reconfigurable Intelligent Surfaces in Indoor Dense Spaces

Contact Info

Product

Resources

About