Reinforcement Learning driven Energy Efficient Mobile Communication and Applications
Summary (2 min read)
I. INTRODUCTION
- This would result in potential rise in energy consumption.
- To mitigate the impacts on the environment with such increased energy consumption, cell switching and traffic offloading is required in an effective manner which would have direct impact on overall operational expenditure, cell power and energy consumptions, and CO 2 emissions.
- As this has been brought into various discussions that a MBS has limited mobile network channels offered by regulatory authority to transmit on a limited scale to serve number of users [4] .
- These challenges lead to a conclusion discussed in many literatures such as [5] , that traditional Macro Cells (MCs) with large coverage footprints would be broken into multiple SCs.
- In Joint traffic offloading, both vertical and horizontal schemes are used.
A. HetNet Architecture
- An approach to densify the network where multiple SCs are deployed under one MC footprint has been proven an effective method to improve capacity.
- This results, with the small coverage radius compared to conventional MC where SCs transmission power is reduced which eventually enhances capacity, reduces cost and improves EE of the network.
- With the discussed approach, several technical challenges start to occur which includes unpremeditated deployment, intercell interference, non-seamless handovers, back-haul overload and inefficient energy consumption.
- The authors main goal, as a first step, is to design a wireless network to derive overall energy consumption, therefore twotier HetNet model is considered.
- Vertical offloading, is a technique to provide continuous service across all SCs within HetNet where user does not experience any transference of services during the offloading procedure.
B. Energy Consumption Model
- For wireless network performance evaluation, the broadly accepted state of the art is to analyse components of RAN at system level.
- There are multiple components in a typical BS that contributes to certain level of power consumption depend on traffic load profiles.
- These components include, power amplifiers, back-haul links, amplifier efficiency, signal processing and generation, air conditioning and others.
- Therefore, from (3), the total energy consumption E HetNet for each time interval t would be determined.
III. PROPOSED METHODOLOGY
- Reinforcement learning driven vertical offloading method proposed in this work uses Q-Learning (QL) algorithm for sequential decision making variant on cell load conditions.
- Due to the low transmit powers of SCs in horizontal offloading and have limitations to a certain range, horizontal offloading can not always be realised between SCs.
- QL algorithm has also proven capability of interacting in dynamic environments [10] with the six main components as (i) agent, (ii) environment, (iii) action, (iv) state, (v) reward/penalty, and (vi) action-value table.
- After the execution of each agent's action, resulting state and reward/penalty are evaluated.
- In different time intervals, MC obtains and records varying traffic condition of SCs in order to make decisions and eventually select set of SCs that are needed to be switched off.
A. Data Set
- This section describes the distribution of users within each cell (either MC or SC) that are used to produce expected capacity over time in HetNet architecture.
- Therefore, by using BS static power, BS transmitted power and dependant load component, total power consumptions of all cells from (4), ( 5) are calculated.
- The overall EE from ( 6) has also been plotted after running 100 iterations and averaging the values.
- The plot is shown in Fig. 3 where the authors have assumed 50% of the subscribers are heavy data users with average data rate of 2 Mb/s multiplied by number of users in each interval.
- There are many ways to calculate user demands by adjusting the ratio of low, medium and heavy users.
B. Benchmarking
- In addition to the proposed Q-learning based CS approach, three more techniques are also developed to compare and assess the performance of the proposed method.
- All the SCs are always kept on, meaning that no switching is implemented, also known as All-On.
- Having this method in the results is quite important, since it is currently the case for the majority of the networks.
- Even though this method does not offer any saving in power consumption and/or CO 2 emission, it does not suffer from reduced quality of service (QoS) given that all the users are kept connected with their best serving BS due to the fact that there is no switching and offloading.
- (b) Gain on the total energy consumption for different methods compared to the All-On method where there is no-switching applied.
C. Metrics
- Three different metrics, namely total energy consumption, percentage gain in the total energy consumption, and the CO 2 emission, are used in this work in order to evaluate the performance of the developed CS techniques.
- In other words, the power consumption of the BSs (either MC or SC) are calculated individually and then combined together to obtain the overall power consumption.
- Lastly, to better reflect the energy saving results to establish the CO 2 footprint on the environment, this study proposes a process for formulating CO 2 emission reduction of overall HetNet architecture when All-On, All-Off, ES and QL methods are envisaged.
- Carbon emissions have been calculated by using conversion factor as shown in (12) .
Did you find this useful? Give us your feedback
Citations
27 citations
Cites background from "Reinforcement Learning driven Energ..."
...Hence, there is a need to develop intelligent traffic prediction and load adaptive cell switching techniques (Feng et al., 2017; Abubakar et al., 2019; Asad et al., 2019), such that the traffic demand on the network can be continually monitored to identify underutilized BSs and automatically switch them off....
[...]
16 citations
Cites background from "Reinforcement Learning driven Energ..."
...However, there is a need for devising mitigation strategies to further decelerate the disease spread by using Artificial Intelligence (AI) [2] that exhibits traits associated with the historical mobility of the different age-groups such as vulnerable age-group travelers....
[...]
11 citations
Cites background from "Reinforcement Learning driven Energ..."
...57] β ∈ [5, 43] β ∈ [9, 38] β ∈ [3, 15] 5....
[...]
...With the limited information, optimisation and encryption would become significantly challenging, which eventually leads to ineffective resource management and a high number of unnecessary deployments that have CO2 emissions [3,5] and costs [6]....
[...]
9 citations
7 citations
References
37,989 citations
"Reinforcement Learning driven Energ..." refers methods in this paper
...Due to its low computational overhead for BS switching QL algorithm proved to be the most chosen solution [15]....
[...]
...In other words, it a method of asynchronous dynamic programming where it provides agents with the opportunity of learning that finds an estimate of the optimal action-value function by experiencing concurrent sequences of actions [15]....
[...]
1,462 citations
587 citations
"Reinforcement Learning driven Energ..." refers background or methods in this paper
...Carbon Emissions Use of carbon footprint (CO2 emissions) is based on total energy of HetNet and can be calculated with the help of conversion factor described in [1], [14]....
[...]
...INTRODUCTION Mobile Communication is responsible for 2% of global CO2 emissions with the potential to increase to approximately 4% by 2020 [1], [2] where data is in high demands likely to increase manifold....
[...]
527 citations
"Reinforcement Learning driven Energ..." refers background in this paper
...[13] that incorporates several radio parameters such as channel capacity, signal strength, signal quality, speed, and transmit power....
[...]
403 citations