Shao-Hung Cheng
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Shao-Hung Cheng has a total of 17 co-authors. Here are the top co-authors by number of shared publications:
- Shu-Chuan Chen
Department of Electrical and Electronic Engineering, Chung Cheng Institute of Technology, National Defense University, Taoyuan, Taiwan
5 shared publications - Li-Chun Wang
Department of Electrical and Computer Engineering National Yang Ming Chiao Tung University
4 shared publications - Chih-Chung Lin
Electrical and Electronic Engineering Department, Chung Cheng Institute of Technology, National Defense University, Taoyuan 335, Taiwan
2 shared publications - Chang-Sheng Wu
Electrical and Electronic Engineering Department, Chung Cheng Institute of Technology, National Defense University, Taoyuan 335, Taiwan
1 shared publication Chang‐Sheng Wu
Department of Electrical and Electronic Engineering Chung Cheng Institute of Technology National Defense University Taoyuan Taiwan
1 shared publication- Cheng-Siang Wei
Electrical Engineering Department, National Yunlin University of Science and Technology, Douliu 640, Taiwan
1 shared publication
Latest Publications
- 9 November 2025
- Published by Wiley in Microwave and Optical Technology Letters
- Vol. 67 (11)
Wideband Square‐Ring Patch Antenna With Coupled Feeds for 5G MIMO Systems
This paper proposes a 5G MIMO wideband four‐coupled‐feed square‐ring patch antenna system specifically designed for access point applications in next‐generation wireless communication networks. The system features a compact and low‐profile form factor, with an overall footprint of 50 × 50 mm² and a height of 10 mm, making it well‐suited for integration into space‐constrained devices. The antenna array is built on a 35 × 35 mm² FR4 substrate with a thickness of 0.8 mm, corresponding to approximately 0.484 × 0.484 λ at the center frequency of 4.15 GHz. Each of the four antenna elements comprises a rectangular metal ring structure with dimensions of 19 × 13 mm² and is symmetrically arranged in a rotational configuration to ensure polarization orthogonality. This design approach effectively enhances port isolation, maintaining port isolation above 10 dB, without the need for additional decoupling structures. The antenna system operates efficiently across a wide frequency band from 3.3 GHz to 5.0 GHz, covering the sub‐6 GHz spectrum allocated for 5G communication. Key MIMO performance indicators, including the envelope correlation coefficient (ECC) and channel capacity loss (CCL), are maintained below 0.09 and 0.2, respectively, across the entire band, signifying low mutual coupling.
- 9 November 2025
- Published by Wiley in Microwave and Optical Technology Letters
- Vol. 67 (11)
Wideband Square‐Ring Patch Antenna With Coupled Feeds for 5G MIMO Systems
This paper proposes a 5G MIMO wideband four‐coupled‐feed square‐ring patch antenna system specifically designed for access point applications in next‐generation wireless communication networks. The system features a compact and low‐profile form factor, with an overall footprint of 50 × 50 mm² and a height of 10 mm, making it well‐suited for integration into space‐constrained devices. The antenna array is built on a 35 × 35 mm² FR4 substrate with a thickness of 0.8 mm, corresponding to approximately 0.484 × 0.484 λ at the center frequency of 4.15 GHz. Each of the four antenna elements comprises a rectangular metal ring structure with dimensions of 19 × 13 mm² and is symmetrically arranged in a rotational configuration to ensure polarization orthogonality. This design approach effectively enhances port isolation, maintaining port isolation above 10 dB, without the need for additional decoupling structures. The antenna system operates efficiently across a wide frequency band from 3.3 GHz to 5.0 GHz, covering the sub‐6 GHz spectrum allocated for 5G communication. Key MIMO performance indicators, including the envelope correlation coefficient (ECC) and channel capacity loss (CCL), are maintained below 0.09 and 0.2, respectively, across the entire band, signifying low mutual coupling.
- 9 December 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Antennas and Wireless Propagation Letters
- Vol. 24 (3) , 696-700
Low-Profile MIMO Trapezoidal Patch Antenna for 5G Wideband Mobile Antenna Application
This letter proposes a low-profile trapezoidal patch quad antenna for 5G broadband mobile applications. The multiple-input multiple-output (MIMO) four-antenna is printed on an FR4 glass fiber substrate with a size of 42 × 42 mm 2 (about 0.581 λ × 0.581 λ at 4.15 GHz) and a thickness of 0.4 mm. The system ground plane is 42 × 42 mm 2 , and the air layer is 1.6 mm above the ground plane. The total cross-sectional height is only 2.0 mm (about 0.027 λ at 4.15 GHz). The antenna features four identically-sized coupled feed structures, each with a 90-degree rotated trapezoidal patch antenna. A shorting pin near the feeding point is added to improve matching and isolation, successfully covering the sub-6-GHz broadband spectrum from 3.3 to 5.0 GHz and achieving a 41% bandwidth at 4.15 GHz. The isolation between antennas is more than 10 dB. The calculated envelope correlation coefficient (ECC) values are less than 0.13. The radiation efficiency of the four antennas has been measured to exceed 40%.
- 9 December 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Antennas and Wireless Propagation Letters
- Vol. 24 (3) , 696-700
Low-Profile MIMO Trapezoidal Patch Antenna for 5G Wideband Mobile Antenna Application
This letter proposes a low-profile trapezoidal patch quad antenna for 5G broadband mobile applications. The multiple-input multiple-output (MIMO) four-antenna is printed on an FR4 glass fiber substrate with a size of 42 × 42 mm 2 (about 0.581 λ × 0.581 λ at 4.15 GHz) and a thickness of 0.4 mm. The system ground plane is 42 × 42 mm 2 , and the air layer is 1.6 mm above the ground plane. The total cross-sectional height is only 2.0 mm (about 0.027 λ at 4.15 GHz). The antenna features four identically-sized coupled feed structures, each with a 90-degree rotated trapezoidal patch antenna. A shorting pin near the feeding point is added to improve matching and isolation, successfully covering the sub-6-GHz broadband spectrum from 3.3 to 5.0 GHz and achieving a 41% bandwidth at 4.15 GHz. The isolation between antennas is more than 10 dB. The calculated envelope correlation coefficient (ECC) values are less than 0.13. The radiation efficiency of the four antennas has been measured to exceed 40%.
- Open Access
- 21 June 2024
- Published by MDPI AG in Electronics
- Vol. 13 (13) , 2430
Compact Asymmetric T-Feed Closed-Slot Antennas for 2.4/5/6 GHz WiFi-7 MIMO Laptops
In this paper, we introduce a closed-slot WiFi-7 multi-input multi-output (MIMO) system equipped with six antennas, designed specifically for laptop integration. Positioned near the lower edge of the laptop’s display ground plane, the antennas are placed 2.5 mm from the hinge and 8 mm from the left side of the ground plane. Each antenna features a 44 × 2 mm2 closed-slot structure with a simple T-shaped feeder that stimulates half-wavelength and full-wavelength resonances via feed-in coupling. The six antennas are aligned in parallel, spaced 4 mm apart, forming a compact array without the need for additional isolation components. This setup supports dual-band functionality across both 2400–2500 MHz and 5150–7125 MHz WiFi-7 bands. Performance assessments indicate a minimum of 10 dB isolation between the antennas and envelope correlation coefficients (ECC) of the radiation patterns being below 0.04. Furthermore, the antenna array’s radiation efficiency was measured to be over 64%.
- 21 June 2024
- Published by MDPI AG in Electronics
- Vol. 13 (13) , 2430
Open AccessCompact Asymmetric T-Feed Closed-Slot Antennas for 2.4/5/6 GHz WiFi-7 MIMO Laptops
In this paper, we introduce a closed-slot WiFi-7 multi-input multi-output (MIMO) system equipped with six antennas, designed specifically for laptop integration. Positioned near the lower edge of the laptop’s display ground plane, the antennas are placed 2.5 mm from the hinge and 8 mm from the left side of the ground plane. Each antenna features a 44 × 2 mm2 closed-slot structure with a simple T-shaped feeder that stimulates half-wavelength and full-wavelength resonances via feed-in coupling. The six antennas are aligned in parallel, spaced 4 mm apart, forming a compact array without the need for additional isolation components. This setup supports dual-band functionality across both 2400–2500 MHz and 5150–7125 MHz WiFi-7 bands. Performance assessments indicate a minimum of 10 dB isolation between the antennas and envelope correlation coefficients (ECC) of the radiation patterns being below 0.04. Furthermore, the antenna array’s radiation efficiency was measured to be over 64%.
- Open Access
- 26 December 2023
- Published by Institution of Engineering and Technology (IET) in IET Microwaves, Antennas & Propagation
- Vol. 18 (2) , 116-127
Application of switchable 4G/5G dual antenna system for laptops
This work investigates a switchable 4G/5G dual‐antenna system designed for laptops. The dual open‐slot antenna structure is straightforward, with a slot width of only 2 mm. Each individual antenna features two different feed sections located in two separate regions between the slots. Port 1 is a direct‐feeding design with 3.6 pF chip capacitance, while Port 2 is only a direct‐feeding design. When Port 1 is excited and Port 2 is terminated to 50‐Ω, the antenna is functional in long‐term evolution (LTE) low‐band (617–960 MHz) and 5G sub‐6 GHz band (3300–5000 MHz). On the other hand, when Port 2 is excited and Port 1 is terminated to 50‐Ω, the antenna is functional in the LTE medium/high bands between 1710 and 2690 MHz. The measured isolation consistently exceeded 30 dB, and the efficiency remained above 48% across the entire frequency range. The calculated envelope correlation coefficients derived from the complex E ‐field radiation pattern measurements were consistently below 0.02. These findings indicate that the system is exceptionally well‐suited for applications characterised by integrated metallic environments and a high display‐to‐body ratio.
- 26 December 2023
- Published by Institution of Engineering and Technology (IET) in IET Microwaves, Antennas & Propagation
- Vol. 18 (2) , 116-127
Open AccessApplication of switchable 4G/5G dual antenna system for laptops
This work investigates a switchable 4G/5G dual‐antenna system designed for laptops. The dual open‐slot antenna structure is straightforward, with a slot width of only 2 mm. Each individual antenna features two different feed sections located in two separate regions between the slots. Port 1 is a direct‐feeding design with 3.6 pF chip capacitance, while Port 2 is only a direct‐feeding design. When Port 1 is excited and Port 2 is terminated to 50‐Ω, the antenna is functional in long‐term evolution (LTE) low‐band (617–960 MHz) and 5G sub‐6 GHz band (3300–5000 MHz). On the other hand, when Port 2 is excited and Port 1 is terminated to 50‐Ω, the antenna is functional in the LTE medium/high bands between 1710 and 2690 MHz. The measured isolation consistently exceeded 30 dB, and the efficiency remained above 48% across the entire frequency range. The calculated envelope correlation coefficients derived from the complex E ‐field radiation pattern measurements were consistently below 0.02. These findings indicate that the system is exceptionally well‐suited for applications characterised by integrated metallic environments and a high display‐to‐body ratio.
- Open Access
- 9 July 2023
- Published by MDPI AG in Micromachines
- Vol. 14 (7) , 1399
Compact Sub 6 GHz Dual Band Twelve-Element MIMO Antenna for 5G Metal-Rimmed Smartphone Applications
In this paper, a twelve-antenna system is designed for 5G smartphones with metal frames. The system is compact and operates on dual bands within the sub-6 GHz frequency range using multiple-input multiple-output (MIMO) technology. Two sets of six-antenna units are included in the system, arranged in a diagonal mirror-image configuration, and positioned at the center of the circuit board’s longer edges. The profile height of each of the six-antenna units is only 3 mm, and the overall array dimensions are 105 × 3 × 3.1 mm3. A single antenna unit is 15 × 3 × 3.1 mm3 (0.173 λ × 0.035 λ × 0.036 λ, where λ equals the free-space wavelength of 3450 MHz). The arrangement of the antennas in the six-antenna units is parallel, with a 3 mm separation between adjacent antennas. The antenna structure comprises of an inverted L-shaped feed branch and two inverted L-shaped short-circuit branches integrated into part of the metal frame. The proposed array can form multiple resonance paths, achieving dual-band operation at 3300–3600 MHz and 4800–5000 MHz. The measured isolation of this twelve-antenna system within the operating frequency band is over 10 dB, and the measured antenna efficiency is greater than 36%. Therefore, the system is suitable for use in smartphones with high screen-to-body ratios and metal frames.
- 9 July 2023
- Published by MDPI AG in Micromachines
- Vol. 14 (7) , 1399
Open AccessCompact Sub 6 GHz Dual Band Twelve-Element MIMO Antenna for 5G Metal-Rimmed Smartphone Applications
In this paper, a twelve-antenna system is designed for 5G smartphones with metal frames. The system is compact and operates on dual bands within the sub-6 GHz frequency range using multiple-input multiple-output (MIMO) technology. Two sets of six-antenna units are included in the system, arranged in a diagonal mirror-image configuration, and positioned at the center of the circuit board’s longer edges. The profile height of each of the six-antenna units is only 3 mm, and the overall array dimensions are 105 × 3 × 3.1 mm3. A single antenna unit is 15 × 3 × 3.1 mm3 (0.173 λ × 0.035 λ × 0.036 λ, where λ equals the free-space wavelength of 3450 MHz). The arrangement of the antennas in the six-antenna units is parallel, with a 3 mm separation between adjacent antennas. The antenna structure comprises of an inverted L-shaped feed branch and two inverted L-shaped short-circuit branches integrated into part of the metal frame. The proposed array can form multiple resonance paths, achieving dual-band operation at 3300–3600 MHz and 4800–5000 MHz. The measured isolation of this twelve-antenna system within the operating frequency band is over 10 dB, and the measured antenna efficiency is greater than 36%. Therefore, the system is suitable for use in smartphones with high screen-to-body ratios and metal frames.
- Open Access
- 9 March 2023
- Published by MDPI AG in Micromachines
- Vol. 14 (3) , 626
Compact Sub 6 GHz Slot Multiantenna System for 5G Laptops
A sub 6 GHz dual-band closed-slot multiantenna system for 5G laptops is proposed in this paper. It was installed in a clearance space, with dimensions og 217 × 3 mm2 and 1 mm away from the upper edge of the screen ground plane. The dimensions of the clearance space were the same as those of a multisystem consisting of six antennas. The dimensions of the single closed-slot antenna were 32 × 3 mm2 (0.368 × 0.034 , where equals the free-space wavelength of 3450 MHz. The antenna was coupled to an asymmetric T-shaped feed-in section equipped with a chip capacitor for exciting one-half and full wavelength resonance modes of the closed-slot to encompass sub 6 GHz 3300–3600 MHz and 4800–5000 MHz dual-band operations. The design of the antenna features a long and straight slot to generate the high-order mode of the closed slot in the high-frequency (4800–5000 MHz) band (not the low-frequency (3300–3600 MHz) multiplier band). Its structure is simple, and the width of its slot is only 3 mm. The antennas were arranged to be 5 mm apart in the same direction and in parallel to form a six-antenna system in order to utilize the weak electric fields located at the two closed ends of the closed-slot structure when the closed slot was excited. It showed excellent envelope correlation coefficients (ECCs) and isolation performance without the installation of isolation elements. The measured fractional bandwidth of the antenna was 10.15% and 6.73% at the center frequencies of 3450 MHz and 4900 MHz, respectively. Its measured isolation was always over 10 dB, and the efficiency was between 46% and 76%. The ECCs of the system calculated from the measured complex E-field radiation pattern were all below 0.2, which means that it is ideal for use in laptop devices with a high screen-to-body ratio and a metal back cover.- 9 March 2023
- Published by MDPI AG in Micromachines
- Vol. 14 (3) , 626
Open AccessCompact Sub 6 GHz Slot Multiantenna System for 5G Laptops
A sub 6 GHz dual-band closed-slot multiantenna system for 5G laptops is proposed in this paper. It was installed in a clearance space, with dimensions og 217 × 3 mm2 and 1 mm away from the upper edge of the screen ground plane. The dimensions of the clearance space were the same as those of a multisystem consisting of six antennas. The dimensions of the single closed-slot antenna were 32 × 3 mm2 (0.368 × 0.034 , where equals the free-space wavelength of 3450 MHz. The antenna was coupled to an asymmetric T-shaped feed-in section equipped with a chip capacitor for exciting one-half and full wavelength resonance modes of the closed-slot to encompass sub 6 GHz 3300–3600 MHz and 4800–5000 MHz dual-band operations. The design of the antenna features a long and straight slot to generate the high-order mode of the closed slot in the high-frequency (4800–5000 MHz) band (not the low-frequency (3300–3600 MHz) multiplier band). Its structure is simple, and the width of its slot is only 3 mm. The antennas were arranged to be 5 mm apart in the same direction and in parallel to form a six-antenna system in order to utilize the weak electric fields located at the two closed ends of the closed-slot structure when the closed slot was excited. It showed excellent envelope correlation coefficients (ECCs) and isolation performance without the installation of isolation elements. The measured fractional bandwidth of the antenna was 10.15% and 6.73% at the center frequencies of 3450 MHz and 4900 MHz, respectively. Its measured isolation was always over 10 dB, and the efficiency was between 46% and 76%. The ECCs of the system calculated from the measured complex E-field radiation pattern were all below 0.2, which means that it is ideal for use in laptop devices with a high screen-to-body ratio and a metal back cover. - Open Access
- 14 September 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Open Journal of Vehicular Technology
- Vol. 2, 412-418
Controlling Interference Structure and Transmit Power of Aerial Small Cells by Hybrid Affinity Propagation Clustering and Reinforcement Learning
This article presents a learning-based interference management mechanism for multiple unmanned aerial vehicles mounted small cells (ASCs), called HAPPIER, standing for hybrid affinity propagation clustering (APC) and reinforcement learning (RL) power control. The proposed HAPPIER interference management mechanism consists of two main algorithms: APC and RL. First, from the macroscopic viewpoint, the APC explores the interference structure of multiple ASCs and then changes the most serious interfering ASCs into sleeping mode. As such, we can shift the complicated interference structure into the one with fewer interfering sources and thus speed up the learning process of interference management. Secondly, from the microscopic viewpoint, based on the interference structure suggested by HAPPIER, the RL is applied to adjust the transmission power of active ASCs to optimize the total throughput further. HAPPIER can achieve the optimal trade-off between system throughput and complexity. From our numerical results, subject to the same complexity constraint, our proposed HAPPIER outperforms all the existing approaches and can achieve 93% of the system throughput of the exhaustive searching algorithm.
- 14 September 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Open Journal of Vehicular Technology
- Vol. 2, 412-418
Open AccessControlling Interference Structure and Transmit Power of Aerial Small Cells by Hybrid Affinity Propagation Clustering and Reinforcement Learning
This article presents a learning-based interference management mechanism for multiple unmanned aerial vehicles mounted small cells (ASCs), called HAPPIER, standing for hybrid affinity propagation clustering (APC) and reinforcement learning (RL) power control. The proposed HAPPIER interference management mechanism consists of two main algorithms: APC and RL. First, from the macroscopic viewpoint, the APC explores the interference structure of multiple ASCs and then changes the most serious interfering ASCs into sleeping mode. As such, we can shift the complicated interference structure into the one with fewer interfering sources and thus speed up the learning process of interference management. Secondly, from the microscopic viewpoint, based on the interference structure suggested by HAPPIER, the RL is applied to adjust the transmission power of active ASCs to optimize the total throughput further. HAPPIER can achieve the optimal trade-off between system throughput and complexity. From our numerical results, subject to the same complexity constraint, our proposed HAPPIER outperforms all the existing approaches and can achieve 93% of the system throughput of the exhaustive searching algorithm.
- Open Access
- 10 September 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 9, 126735-126747
Proactive Power Control and Position Deployment for Drone Small Cells: Joint Supervised and Unsupervised Learning
Since unmanned aerial vehicles (UAV) are easily deployed, highly mobile, and hover capability, they are utilized for many commercial applications. In particular, small cells mounted on UAVs, also known as drone small cells (DSC), may provide temporary relief or ancillary programs for the wireless network. In this paper, we design a prediction-based proactive drone management (P 2 DM) framework to reduce network interference and improve energy efficiency in the multiple DSCs scenario. The P 2 DM framework can be divided into offline and online phases. In the offline phase, supervised learning is used to build a highly accurate mobility prediction model according to the historical data. The prediction model is launched in the online phase to predict the user position only using a small sample set. The system proactively determines whether a DSC should be awake or asleep at the next timeslot due to the predicted user positions. Since DSC has more longer awake time in the deep sleeping mode, it is previously awoken to avoid data propagation delay. To further overcome the difficulty of obtaining the key performance indicator data (i.e., labeled data) in the online phase, an unsupervised learning technique is employed for DSC repositioning and power control to improve energy efficiency. Our simulation results show that the P 2 DM framework can demonstrate the advantage in terms of execution time and energy efficiency compared to the existing method based on genetic algorithm (i.e., a heuristic algorithm).
- 10 September 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 9, 126735-126747
Open AccessProactive Power Control and Position Deployment for Drone Small Cells: Joint Supervised and Unsupervised Learning
Since unmanned aerial vehicles (UAV) are easily deployed, highly mobile, and hover capability, they are utilized for many commercial applications. In particular, small cells mounted on UAVs, also known as drone small cells (DSC), may provide temporary relief or ancillary programs for the wireless network. In this paper, we design a prediction-based proactive drone management (P 2 DM) framework to reduce network interference and improve energy efficiency in the multiple DSCs scenario. The P 2 DM framework can be divided into offline and online phases. In the offline phase, supervised learning is used to build a highly accurate mobility prediction model according to the historical data. The prediction model is launched in the online phase to predict the user position only using a small sample set. The system proactively determines whether a DSC should be awake or asleep at the next timeslot due to the predicted user positions. Since DSC has more longer awake time in the deep sleeping mode, it is previously awoken to avoid data propagation delay. To further overcome the difficulty of obtaining the key performance indicator data (i.e., labeled data) in the online phase, an unsupervised learning technique is employed for DSC repositioning and power control to improve energy efficiency. Our simulation results show that the P 2 DM framework can demonstrate the advantage in terms of execution time and energy efficiency compared to the existing method based on genetic algorithm (i.e., a heuristic algorithm).
- Open Access
- 11 October 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Cognitive Communications and Networking
- Vol. 6 (1) , 83-94
An Integrated Affinity Propagation and Machine Learning Approach for Interference Management in Drone Base Stations
- 11 October 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Cognitive Communications and Networking
- Vol. 6 (1) , 83-94
Open AccessAn Integrated Affinity Propagation and Machine Learning Approach for Interference Management in Drone Base Stations
- Open Access
- 16 July 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Systems Journal
- Vol. 13 (2) , 1397-1408
Self-Organizing Ultra-Dense Small Cells in Dynamic Environments: A Data-Driven Approach
This paper presents a data-driven biadaptive self-organizing network (Bi-SON) for ultra-dense small cells (UDSC), which can improve energy efficiency and reduce interference in dynamic environments, taking account of cell switching on/off, transmission power adjustment, and traffic loads simultaneously. In the first adaptation of Bi-SON, a joint traffic load and interference aware cell ranking mechanism first determines the necessary active small cells based on traffic loads, and then ranks all the active small cells based on their carried traffic load and resulting interference. Top ranked cells will transmit at the maximum power. The last ranked $K$ cells will adjust the transmission power for interference reduction in the second adaptation function of Bi-SON, while maintaining the required quality of service. According to a polynomial regression learning approach, the total system throughput of UDSC is characterized as a function of $K$ . Compared to the baseline case when all the cells transmit with the maximum power, our proposed Bi-SON framework can improve the throughput and energy efficiency of UDSC by 73% and 169%, respectively. However, the pure switching on/off approach can only improve the throughput and the energy efficiency of UDSC by 52% and 115%, respectively. As demonstrated, even with a simple power adaptation algorithm, a learning-based Bi-SON framework can improve the performance of UDSC by taking advantage of the pervasive availability of voluminous data.- 16 July 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Systems Journal
- Vol. 13 (2) , 1397-1408
Open AccessSelf-Organizing Ultra-Dense Small Cells in Dynamic Environments: A Data-Driven Approach
This paper presents a data-driven biadaptive self-organizing network (Bi-SON) for ultra-dense small cells (UDSC), which can improve energy efficiency and reduce interference in dynamic environments, taking account of cell switching on/off, transmission power adjustment, and traffic loads simultaneously. In the first adaptation of Bi-SON, a joint traffic load and interference aware cell ranking mechanism first determines the necessary active small cells based on traffic loads, and then ranks all the active small cells based on their carried traffic load and resulting interference. Top ranked cells will transmit at the maximum power. The last ranked $K$ cells will adjust the transmission power for interference reduction in the second adaptation function of Bi-SON, while maintaining the required quality of service. According to a polynomial regression learning approach, the total system throughput of UDSC is characterized as a function of $K$ . Compared to the baseline case when all the cells transmit with the maximum power, our proposed Bi-SON framework can improve the throughput and energy efficiency of UDSC by 73% and 169%, respectively. However, the pure switching on/off approach can only improve the throughput and the energy efficiency of UDSC by 52% and 115%, respectively. As demonstrated, even with a simple power adaptation algorithm, a learning-based Bi-SON framework can improve the performance of UDSC by taking advantage of the pervasive availability of voluminous data.