# A Novel Interference Avoidance Based on a Distributed Deep Learning Model for 5G-Enabled IoT

^{1}

^{2}

^{3}

^{*}

## Abstract

**:**

## 1. Introduction

- The proposed approach developed an efficient method to enhance the overall system performance in terms of system throughput and energy efficiency.
- An optimization problem using an analytical and deep learning model was formulated to ascertain the reliability and efficiency of communication among 5G and IoTs.
- The proposed approach aims to decrease or eliminate the interference in 5G networks and IoT systems. This was achieved through determining the optimum distance between CUE-IoTG and IoTD-D for the uplink (UL) data communication and between BS-IoTD and IoTG-CUE for the downlink (DL) data communication. This can be achieved based on different parameters, which affect the system performance such as transmission power, distance between CUE-D and IoTD-IoTG, path loss and signal-to-interference-plus-noise ratio (SINR
_{th}). - The proposed approach allowed the transmission of CUE and IoTD, using a deep learning model, to predict the suitable acceptable distance between CUE-IoTG and IoTD-D (uplink) and between BS-IoTD and IoTG-CUE (downlink) thus avoiding severe interference.
- The proposed deep learning model was compared to state-of-the-art benchmark methods and it provided a marked improvement in the results.
- The proposed model can be used in the design phase for interference prediction and circumvention.
- The proposed approach was investigated in terms of overall system throughput and energy efficiency under different conditions, such as the path loss exponent, transmission power, different SINR
_{th}values, and different transmission ranges. The whole network can be optimized by these findings in a vibrant environment.

## 2. Related Work

## 3. Proposed Model

#### 3.1. System Model and Problem Formulation

#### 3.1.1. Uplink Data Communication

_{,}${\gamma}_{IG}$, ${\gamma}_{CiG}$

_{,}and ${\gamma}_{IjD}$ represent the path loss model of CUE-D, IoTD-IoTG, i-th CUE-IoTG and the j-th IoTD-D, respectively. The path loss between CUE-D, IoTD-IoTG, i-th CUE-IoTG and the j-th IoTD-D can be expressed as [29,30]:

#### 3.1.2. Downlink Data Communication

_{,}${\gamma}_{GC}$, ${\gamma}_{GI}$

_{,}and ${\gamma}_{BI}$ represent the path loss model of BS-CUE, IoTG-CUE, IoTG-IoTD and BS-IoTD, respectively. Thus, the path loss between BS-CUE, IoTG-CUE, IoTG-IoTD and BS-IoTD can be expressed as [29,30]:

_{thCiB}

_{,}SINR

_{thBCi}) and IoT system (SINR

_{thIjG}, SINR

_{thGIj}) has the same value, which is SINR

_{th}

_{.}

#### 3.2. Dataset Generation

_{th}and distances of CUE-D and IoTD-IoTG using the Lagrange optimization technique to generate the optimal distances of IoTD-D and CUE-IoTG for each input. The experiments were run for different values of SINR

_{th}ranging from 0 to 20. For each value of SINR

_{th}, the value of the CUE-D distance was initialized to 1 and incremented by half a meter for each record until the throughput and energy efficiency of the calculated distances of IoTD-D and CUE-IoTG were unacceptable.

_{th}

_{,}along with distances of BS-CUE and IoTG-IoTD. The output distances were evaluated for each record to make sure that they meet the required throughput and energy efficiency.

#### 3.3. Proposed Deep Learning Model

#### 3.3.1. Network Structure

_{th}, input distance 1 (I-Dist1) and input distance 2 (I-Dist2). I-Dist1 represent the CUE-D for uplink communication and BS-CUE for downlink communication while I-Dist2 denotes IoTD-IoTG for uplink communication and IoTG-IoTD for communication.

- An abstract input layer that takes the current values of the input and passes it to the 1D-CNN layers
- The first 1D-CNN is 3 × 1 having 32 filters, with a kernel size of 3
- The second 1D-CNN is 1 × 1 having 16 filters, with a kernel size of 1
- A flattening layer to reshape the 1D CNN can be input to the fully connected layers
- A 32-neuron fully connected layer
- A 16-neuron fully connected layer
- An output layer to produce the regression distance result

#### 3.3.2. Data Scaling

#### 3.3.3. Activation Function

#### 3.3.4. Optimization Function

#### 3.3.5. Parameter Optimization

## 4. Results and Discussion

#### 4.1. Deep Learning Model Results Evaluation

- Mean Absolute Error (MAE), which measures the average differences between actual and predicted values.

- Root Mean Squared Error, which calculates the square root of the average of the squared differences between actual and predicted values as

#### 4.2. Analytical Evaluation

_{th}for the proposed model using the analytical and deep learning model. For the uplink data communication, it was assumed that all the transmitted devices, whether they were CUEs or IoTDs, always had a maximum transmission power equal to 23 dBm. Thus, it can be noticed from Figure 5a, for the analytical and deep learning model, the optimum required distance between IoTD-D (d

_{ID}) to decrease the interference at the destination increased when the distance between CUE-D (d

_{CD}) increased. In addition, it can be mentioned that in order for d

_{CD}to reach the maximum value, SINR

_{th}decreased gradually since increasing the transmission distance led to increasing the losses in the communication link. Consequently, decreasing transmission distance increased SINR

_{th}—for example, to have a communication link with SINR

_{th}equal to 0 dB the transmission distance d

_{CD}remained effective until it exceeded 836 m. Additionally, when d

_{CD}was 600.5 m, d

_{ID}must be greater or equal to 647.33 m analytically and 646.8 m using the deep learning model. On the other hand, when the required SINR

_{th}for any communication link was 20 dB, the maximum transmission distance for reaching effective communication was 261.5 m. It can also be noticed that, the required distance between IoTD-D was equal to 317.2 and 316.72 assuming that d

_{CD}was 99.64 m, using analytical and deep learning model, respectively.

_{th}led to an increase in the required distance between IoTG-CUE to avoid interference. For example, when SINR

_{th}was 0 dB and the downlink transmission distance (d

_{BC}) was 600.5 m, the required distance between IoTG-CUE (d

_{GC}) to avoid interference was 505.52 m numerically and 506.95 using deep learning, while when SINR

_{th}was 5 dB, d

_{GC}must be 677.5 m and 677.8 m for numerical and deep learning, respectively. On the other hand, when SINR

_{th}was 20 dB and d

_{BC}was 99.5 it can be found that d

_{GC}must be in the range of 265 m to avoid interference.

_{CG}) and for downlink (d

_{GC}) must be greater than the distance between IoTD and IoTG for uplink (d

_{IG}) and for downlink (d

_{GI}) —for example, as shown in Figure 6a in the case of uplink, when dIG was equal to 240.2 m; the distance dCG should be 240.6 m analytically and 241 m using deep learning when the required SINRth is 0 dB. While, when dIG was equal to 44 m and SINR

_{th}was 20 dB, d

_{CG}was 139.17 m analytically and 139.63 m using deep learning. On the other hand, in case of the downlink as demonstrated in Figure 6b, when the required SINRth was 0 dB and dGI was equal to 240.2 m the distance d

_{GC}should be 285.48 m analytically and 286.35 m using deep learning. While, when d

_{GI}was equal to 44 m, d

_{GC}should be 165.37 m analytically and 165.33 m using deep learning when SINR

_{th}was 20 dB. Additionally, it can be noticed that decreasing the distance between any source and destination links leads to decreasing the path loss and increasing the SINR

_{th}. Based on the proposed model, it is assumed that the transmission distance between CUE-BS (uplink) and BS-CUE (downlink) is greater than the transmission distance between IoTD-IoTG (uplink) and IoTG-IoTD (downlink). It has been concluded from Figure 6 that, the interference distance (d

_{GC}) during the downlink must be greater than the interference distance (d

_{CG}) during the uplink, this due to the highest transmission power of IoTG and the nearest distance between IoTG-IoTD, which increases the interference at CUE. That is why the interference distance during the downlink (d

_{GC}) should increase compared with the interference distance during the uplink (d

_{GC}).

_{ID}and d

_{GC}distance for decreasing the interference at any D (uplink) and decreasing the interference at CUE (downlink) was examined again in Figure 7 for the analytical and deep learning model for different transmission distances and against SINR

_{th}. Different uplink (d

_{CD}) and downlink transmission (d

_{BC}) distance values were assumed such as 66, 140, and 260 m and against SINR

_{th}, which varied from 0 to 20 dB to predict d

_{ID}and d

_{BI}. As can be observed, when SINR

_{th}increases the predicted required uplink d

_{ID}and downlink d

_{BI}must be greater than or equal to uplink d

_{CD}and downlink d

_{BC}, respectively, for decreasing the interference and at the same time satisfying the system requirements in term of SINR

_{th}—for example, as shown in Figure 7a when d

_{CD}was 260 m, the optimum required d

_{ID}was 260.6 m analytically and 264.24 using deep learning when SINR

_{th}= 0 dB, while when SINR

_{th}= 18 dB for the same distance d

_{CD}, the predicted required distance d

_{ID}was 907.72 m analytically and 903.74 using deep learning. On the other hand, when d

_{ID}was 66 m and SINR

_{th}= 4 dB, d

_{CD}must be 83.09 and 85.75 analytically and based on deep learning, respectively. Furthermore, for the downlink data communication as shown in Figure 7b, for SINR

_{th}= 0 dB and d

_{BC}= 260 m, the required distance to avoid interference d

_{GC}was 218.77m and 219.24 using analytical and the deep learning algorithm, respectively, while for the same transmission distance d

_{GC}should be in the range of 619 m to avoid interference and fulfil the required SINR

_{th}, which is 18 dB. Additionally, when d

_{BC}was 66 m, the required distance for avoiding interference should be in the range of 55.6 m if SINR

_{th}is 0 dB and 176 m when SINR

_{th}= 20 dB.

_{CG}and d

_{BI}) for decreasing the interference at IoTG (uplink) and IoTD (downlink), respectively. A different scenario is proposed to evaluate the system performance for the uplink and downlink data communication, assuming that d

_{IG}and d

_{GI}are 104 m, 56 m, and 26.4 m for different values of SINR

_{th}. As shown in Figure 8a, in case of uplink, when the required SINR

_{th}increased, for different transmission d

_{IG}, the distance between d

_{CG}increased—for example, when d

_{IG}was equal to 104 m and SINR

_{th}was equal to 7 dB, the optimum required d

_{CG}for the analytical and deep learning model was 155.65 m and 155.54 m, respectively. However, for the same transmission distance when SINR

_{th}was equal to 18 dB, d

_{CG}was 294.2 m analytically and 294.46 m using the deep learning model. The same performance was obtained when the system was evaluated during the downlink as shown in Figure 8b, e.g., when SINR

_{th}was equal to 7 dB and d

_{GI}was equal to 104 m and, the optimum required d

_{CG}for the analytical and deep learning model was 185 m. However, for the same transmission distance d

_{CG}is 348 m and 348.86 m for the analytically and deep learning model, respectively, when SINR

_{th}was equal to 18 dB. Figure 7 and Figure 8 show that the required SINR

_{th}is an important parameter that should affect the predicted required interference distance for decreasing the interference at any destination. Additionally, using these results and based on the system requirements and environmental conditions, an adaptive smart system should be engaged to enhance the system performance for both CUE and IoT networks.

_{th}values did not exist in testing records. Figure 9 demonstrates the optimized system throughput for the proposed approach for the uplink and downlink data communication using the analytical and deep learning model for different randomly chosen SINR

_{th}values. For fair performance evaluation three different scenarios were considered. Assuming that the distance combinations between the uplink distances (d

_{CD}and d

_{IG}) and the downlink distances (d

_{BC}and d

_{GI}) were considered to be the same; they were 260 m and 104 m, 140 m and 56 m, and 66 m and 26.4 m, respectively. The chosen distances represented long, intermediate, and short distances. As observed for the uplink and downlink data communication represented in Figure 9a,b, respectively, for the three different scenarios when the SINR

_{th}increased, the optimized system throughput increased. Additionally, it can be noticed that for the three scenarios for any SINR

_{th}, the optimized system throughput value was approximately identical for the analytical and deep learning models. This means that the proposed approach is capable of reaching the maximum system throughput regard less of the transmission was (long-intermediate-short), as the aim of the proposed model is to predict the interference transmission distance between any interfering node and any destination, for trying to prevent interference and increase system reliability.

_{th}for the analytical and deep learning models. The proposed model succeeded in keeping the optimized energy efficiency approximately the same for the three different assumed transmission distances. Figure 10 is correlated with the results obtained in Figure 9. These two results show the effectiveness of the proposed model in predicting the position of the interference nodes, as by knowing the distance between them and any destination helps prevent the interference, thus increasing the system performance.

_{th}values with different transmission powers for CUE (P

_{C}) and IoTD (P

_{I}). It was assumed that the values of SINR

_{th}were 5, 10, 15 and 20 dB, respectively. As depicted in Figure 11, for any SINR

_{th}increasing the transmission power leads to increasing the system throughput for the analytical and deep learning model. As the system is always limited by channel noise, pathloss and interference that is why the transmission power is one of the parameters, which can overcome the channel conditions. Thus, increasing or decreasing the transmission power must be considered according to the channel conditions and the required system QoS. Furthermore, by comparing the four different SINR

_{th}values it can be found that increasing the SINR

_{th}increases the overall system throughput, which is correlated with the results obtained in Figure 9.

_{th}stated in Figure 9 and with different P

_{C}and P

_{I}. As shown in Figure 12 each value of SINR

_{th}yields a maximum transmission power that leads to an optimum energy efficiency—for example, when SINR

_{th}= 0 dB, the maximum transmission power for any sender node to reach the optimum energy efficiency is 2 or 4 dBm, while when SINR

_{th}= 5 dB, the maximum power is 4 or 6 dBm to reach the maximum energy efficiency. On the other hand, when SINR

_{th}is 20 dB, the maximum transmission power is energy efficiency is 8 dBm. It can be deduced from this figure that increasing the transmission power may lead to a decrease in the energy efficiency as the increment of the transmission power incr11eases the system cost and decreases the system energy efficiency. By comparing Figure 11 and Figure 12, increasing the transmission power increases the overall system throughput and at the same time may decreases the energy efficiency. Thus, for obtaining the maximum system throughput with the highest energy efficiency, the two performances can be jointly considered in order to obtain the required system performance based on the two metrics.

## 5. Conclusions

_{th}and the transmission power on predicting the maximum required interference distance was investigated. It was shown that increasing SINR

_{th}leads to increasing the interference distance between CUE-IoTG, IoTD-D, IoTG-CUE and BS-IoTD. Moreover, it has been shown that increasing the transmission power increases the overall system performance. Additionally, among different values of transmission power, one can reach the maximum energy efficiency. The obtained results show that the proposed model can achieve the maximum system throughput and energy efficiency with suitable system reliability.

## Author Contributions

## Funding

## Institutional Review Board Statement

## Informed Consent Statement

## Data Availability Statement

## Conflicts of Interest

## References

- Wang, H.; Xiong, D.; Wang, P.; Liu, Y. A Lightweight XMPP Publish/Subscribe Scheme for Resource-Constrained IoT Devices. IEEE Access
**2017**, 5, 16393–16405. [Google Scholar] [CrossRef] - Ahmad, I.; Kumar, T.; Liyanage, M.; Okwuibe, J.; Ylianttila, M.; Gurtov, A. Overview of 5G Security Challenges and Solutions. IEEE Commun. Stand. Mag.
**2018**, 2, 36–43. [Google Scholar] [CrossRef] [Green Version] - Wazid, M.; Das, A.K.; Shetty, S.; Gope, P.; Rodrigues, J.J.P.C. Security in 5G-Enabled Internet of Things Communication: Issues, Challenges, and Future Research Roadmap. IEEE Access
**2021**, 9, 4466–4489. [Google Scholar] [CrossRef] - Luong, N.C.; Hoang, D.T.; Wang, P.; Niyato, D.; Kim, D.I.; Han, Z. Data Collection and Wireless Communication in Internet of Things (IoT) Using Economic Analysis and Pricing Models: A Survey. IEEE Commun. Surv. Tutorials
**2016**, 18, 2546–2590. [Google Scholar] [CrossRef] [Green Version] - Chettri, L.; Bera, R. A Comprehensive Survey on Internet of Things (IoT) Toward 5G Wireless Systems. IEEE Internet Things J.
**2020**, 7, 16–32. [Google Scholar] [CrossRef] - Alhajri, M.I.; Ali, N.T.; Shubair, R.M. Classification of Indoor Environments for IoT Applications: A Machine Learning Approach. IEEE Antennas Wirel. Propag. Lett.
**2018**, 17, 2164–2168. [Google Scholar] [CrossRef] - Abbasi, M.; Shahraki, A.; Taherkordi, A. Deep Learning for Network Traffic Monitoring and Analysis (NTMA): A Survey. Comput. Commun.
**2021**, 170, 19–41. [Google Scholar] [CrossRef] - Hansen, E.B.; Bøgh, S. Artificial intelligence and internet of things in small and medium-sized enterprises: A survey. J. Manuf. Syst.
**2021**, 58, 362–372. [Google Scholar] [CrossRef] - Hattab, G.; Cabric, D. Energy-Efficient Massive IoT Shared Spectrum Access Over UAV-Enabled Cellular Networks. IEEE Trans. Commun.
**2020**, 68, 5633–5648. [Google Scholar] [CrossRef] - Seo, J.-B.; Jung, B.C.; Jin, H. Online Backoff Control for NOMA-Enabled Random Access Procedure for Cellular Networks. IEEE Wirel. Commun. Lett.
**2021**, 10, 1158–1162. [Google Scholar] [CrossRef] - Liu, Y.; Liu, K.; Han, J.; Zhu, L.; Xiao, Z.; Xia, X.-G. Resource Allocation and 3-D Placement for UAV-Enabled Energy-Efficient IoT Communications. IEEE Internet Things J.
**2021**, 8, 1322–1333. [Google Scholar] [CrossRef] - Kolawole, O.Y.; Biswas, S.; Singh, K.; Ratnarajah, T. Transceiver Design for Energy-Efficiency Maximization in mmWave MIMO IoT Networks. IEEE Trans. Green Commun. Netw.
**2019**, 4, 109–123. [Google Scholar] [CrossRef] - Khodakhah, F.; Mahmood, A.; Österberg, P.; Gidlund, M. Multiple Access-Enabled Relaying with Piece-Wise and Forward NOMA: Rate Optimization under Reliability Constraints. Sensors
**2021**, 21, 4783. [Google Scholar] [CrossRef] [PubMed] - Osman, R.A.; Zaki, A.I. Energy-Efficient and Reliable Internet of Things for 5G: A Framework for Interference Control. Electron.
**2020**, 9, 2165. [Google Scholar] [CrossRef] - Budhiraja, I.; Kumar, N.; Tyagi, S. Deep-Reinforcement-Learning-Based Proportional Fair Scheduling Control Scheme for Underlay D2D Communication. IEEE Internet Things J.
**2021**, 8, 3143–3156. [Google Scholar] [CrossRef] - Sakib, S.; Tazrin, T.; Fouda, M.M.; Fadlullah, Z.M.; Nasser, N. An Efficient and Lightweight Predictive Channel Assignment Scheme for Multiband B5G-Enabled Massive IoT: A Deep Learning Approach. IEEE Internet Things J.
**2021**, 8, 5285–5297. [Google Scholar] [CrossRef] - Xiao, L.; Zhang, H.; Xiao, Y.; Wan, X.; Liu, S.; Wang, L.-C.; Poor, H.V. Reinforcement Learning-Based Downlink Interference Control for Ultra-Dense Small Cells. IEEE Trans. Wirel. Commun.
**2019**, 19, 423–434. [Google Scholar] [CrossRef] - Gu, B.; Zhang, X.; Lin, Z.; Alazab, M. Deep Multiagent Reinforcement-Learning-Based Resource Allocation for Internet of Controllable Things. IEEE Internet Things J.
**2021**, 8, 3066–3074. [Google Scholar] [CrossRef] - Kim, J.; Park, J.; Noh, J.; Cho, S. Autonomous Power Allocation Based on Distributed Deep Learning for Device-to-Device Communication Underlaying Cellular Network. IEEE Access
**2020**, 8, 107853–107864. [Google Scholar] [CrossRef] - Xiao, Y.; Niu, G.; Xiao, L.; Ding, Y.; Liu, S.; Fan, Y. Reinforcement learning based energy-efficient internet-of-things video transmission. Intell. Converg. Netw.
**2020**, 1, 258–270. [Google Scholar] [CrossRef] - Azari, A.; Masoudi, M. Interference management for coexisting Internet of Things networks over unlicensed spectrum. Ad Hoc Networks
**2021**, 120, 102539. [Google Scholar] [CrossRef] - Babich, F.; Buttazzoni, G.; Vatta, F.; Comisso, M. Energy-Constrained Design of Joint NOMA-Diversity Schemes with Imperfect Interference Cancellation. Sensors
**2021**, 21, 4194. [Google Scholar] [CrossRef] - Alexandre, L.C.; De Souza Filho, A.L.; Sodré, A.C. Indoor Coexistence Analysis Among 5G New Radio, LTE-A and NB-IoT in the 700 MHz Band. IEEE Access
**2020**, 8, 135000–135010. [Google Scholar] [CrossRef] - Li, Y.; Liang, Y.; Liu, Q.; Wang, H. Resources Allocation in Multicell D2D Communications for Internet of Things. IEEE Internet Things J.
**2018**, 5, 4100–4108. [Google Scholar] [CrossRef] - Fu, S.; Su, Z.; Jia, Y.; Zhou, H.; Jin, Y.; Ren, J.; Wu, B.; Huq, K.M.S. Interference Cooperation via Distributed Game in 5G Networks. IEEE Internet Things J.
**2017**, 6, 311–320. [Google Scholar] [CrossRef] - Siddiqui, M.U.A.; Qamar, F.; Ahmed, F.; Nguyen, Q.N.; Hassan, R. Interference Management in 5G and Beyond Network: Requirements, Challenges and Future Directions. IEEE Access
**2021**, 9, 68932–68965. [Google Scholar] [CrossRef] - Zhang, Q.; Zhang, L.; Liang, Y.-C.; Kam, P.-Y. Backscatter-NOMA: A Symbiotic System of Cellular and Internet-of-Things Networks. IEEE Access
**2019**, 7, 20000–20013. [Google Scholar] [CrossRef] - Abrardo, A.; Moretti, M. Distributed Power Allocation for D2D Communications Underlaying/Overlaying OFDMA Cellular Networks. IEEE Trans. Wirel. Commun.
**2017**, 16, 1466–1479. [Google Scholar] [CrossRef] [Green Version] - Fan, B.; Tian, H.; Jiang, L.; Vasilakos, A.V. A Social-Aware Virtual MAC Protocol for Energy-Efficient D2D Communications Underlying Heterogeneous Cellular Networks. IEEE Trans. Veh. Technol.
**2018**, 67, 8372–8385. [Google Scholar] [CrossRef] - Elhalawany, B.M.; Ruby, R.; Wu, K. D2D Communication for Enabling Internet-of-Things: Outage Probability Analysis. IEEE Trans. Veh. Technol.
**2019**, 68, 2332–2345. [Google Scholar] [CrossRef] - Huang, Y.; Liu, M.; Liu, Y. Energy-Efficient SWIPT in IoT Distributed Antenna Systems. IEEE Internet Things J.
**2018**, 5, 2646–2656. [Google Scholar] [CrossRef] [Green Version] - Chae, S.H.; Jeon, S.-W.; Jeong, C. Efficient Resource Allocation for IoT Cellular Networks in the Presence of Inter-Band Interference. IEEE Trans. Commun.
**2019**, 67, 4299–4308. [Google Scholar] [CrossRef] - Liu, C.-H.; Shen, Y.-H.; Lee, C.-H. Energy-Efficient Activation and Uplink Transmission for Cellular IoT. IEEE Internet Things J.
**2020**, 7, 906–921. [Google Scholar] [CrossRef] [Green Version] - Staniec, K.; Kucharzak, M.; Jóskiewicz, Z.; Chowański, B. Measurement-Based Investigations of the NB-IoT Downlink Performance in Fading Channels. IEEE Wirel. Commun. Lett.
**2021**, 10, 1780–1784. [Google Scholar] [CrossRef] - Kiranyaz, S.; Ince, T.; Hamila, R.; Gabbouj, M. Convolutional Neural Networks for patient-specific ECG classification. In Proceedings of the 2015 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), Milan, Italy, 25–29 August 2015; pp. 2608–2611. [Google Scholar]
- Kiranyaz, S.; Avci, O.; Abdeljaber, O.; Ince, T.; Gabbouj, M.; Inman, D.J. 1D convolutional neural networks and applications: A survey. Mech. Syst. Signal Process.
**2021**, 151, 107398. [Google Scholar] [CrossRef] - He, K.; Zhang, X.; Ren, S.; Sun, J. Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification. In Proceedings of the International Conference on Computer Vision (ICCV), Las Condes, Chile, 11–18 December 2015; pp. 1026–1034. [Google Scholar] [CrossRef] [Green Version]
- Kingma, D.P.; Ba, J. Adam: A method for stochastic optimization. arXiv
**2014**, arXiv:1412.6980. [Google Scholar]

**Figure 1.**Proposed network schematic. (

**a**) Uplink data communication (

**b**) Downlink data communication.

**Figure 2.**Spearman correlation of all features generated for (

**a**) the uplink dataset and (

**b**) the downlink dataset.

**Figure 3.**Proposed deep learning network that is to be used by each device independently to calculate optimal output distances.

**Figure 4.**Mean absolute error generated by the training and validation data when calculating the output distances O-Dist1 and O-Dist2 for: (

**a**) uplink communication dataset and (

**b**) downlink communication dataset.

**Figure 5.**(

**a**) Distance between CUE-D versus predicted distance between IoTD-D (uplink) (

**b**) Distance between BS-CUE versus predicted distance between IoTG-CUE (downlink).

**Figure 6.**(

**a**) Distance between IoTD-IoTG versus predicted distance between CUE-IoTG (uplink) (

**b**) Distance between IoTG-IoTD versus predicted distance between BS-IoTD.

**Figure 7.**(

**a**) Signal-to-interference-ratio- plus-noise (SINR

_{th}) versus predicted distance between IoTD-D (uplink) (

**b**) Signal-to-interference-ratio- plus-noise (SINR

_{th}) versus predicted distance between IoTG-CUE (dowlink).

**Figure 8.**(

**a**) Signal-to-interference-ratio-plus-noise (SINR

_{th}) versus predicted distance between CUE-IoTG (uplink) (

**b**) Signal-to-interference-ratio-plus-noise (SINR

_{th}) versus predicted distance between BS-IoTD (downlink).

**Figure 9.**Signal-to-interference-ratio-plus-noise (SINR

_{th}) versus optimized system throughput (

**a**) uplink (

**b**) downlink.

**Figure 10.**Signal-to-interference-ratio-plus-noise (SINR

_{th}) versus optimized energy efficiency (

**a**) uplink (

**b**) downlink.

Parameters | Value |
---|---|

${N}_{o}$ | 174 dBm [31] |

B | 10 MHz |

SINR_{th} | 20 dB [32] |

P_{c} | 23 dBm [32] |

P_{I} | 23 dBm [32] |

P_{B} | 46 dBm [9,25] |

P_{G} | 43 dBm [33,34] |

α | 4 |

γ_{o} | 10^{−1} [32] |

f_{c} | 2 GHz |

SINR_{th} | CUE-D | IoTD-IoTG | IoTD-D | CUE-IoTG | |
---|---|---|---|---|---|

Number of records | 21,055 | 21,055 | 21,055 | 21,055 | 21,055 |

Minimum | 0.00 | 1.00 | 0.40 | 1.00 | 0.40 |

Maximum | 20.00 | 840.00 | 336.00 | 4644.00 | 338.65 |

Mean | 7.94 | 281.23 | 112.49 | 501.97 | 168.77 |

Standard Deviation | 5.84 | 190.82 | 76.33 | 395.39 | 97.39 |

SINR_{th} | BS-CUE | IoTG-IoTD | IoTG-CUE | BS-IoTD | |
---|---|---|---|---|---|

Number of records | 21,055 | 21,055 | 21,055 | 21,055 | 21,055 |

Minimum | 0 | 1 | 0.4 | 0.84 | 0.48 |

Maximum | 20 | 840 | 336 | 709 | 400 |

Mean | 7.94 | 281.23 | 112.49 | 354.39 | 200.15 |

Standard Deviation | 5.84 | 190.82 | 76.33 | 204.09 | 115.22 |

Benchmarks | IoTG-CUE | BS-IoTD |
---|---|---|

Support vector regressor | kernel = ‘rbf’, C = 220, gamma = 40 | Kernel = ‘rbf’, C = 200, gamma = 50 |

Random forest regressor | max_depth = 100, max_features = 3, min_samples_leaf = 3, min_samples_split = 8, n_estimators = 1000 | max_depth = 90, max_features = 3, min_samples_leaf = 3, min_samples_split = 8, n_estimators = 1000 |

Adaboost regressor | learning_rate = 0.01, loss = ‘Linear’, n_estimators = 150 | learning_rate = 1, loss = ‘linear’, n_estimators = 150 |

Multilayer perceptron | activation = ‘tanh’, alpha = 0.05, solver = ‘sgd’, hidden_layer_sizes = (300,), learning_rate = ‘adaptive’ | activation = ‘tanh’, alpha = 0.05, solver = ‘sgd’, hidden_layer_sizes = (300,), learning_rate = ‘adaptive’ |

Benchmarks | IoTG-CUE | BS-IoTD |
---|---|---|

Support vector regressor | kernel = ‘rbf’, C = 220, gamma = 40 | Kernel = ‘rbf’, C = 200, gamma = 50 |

Random forest regressor | max_depth = 100, max_features = 3, min_samples_leaf = 3, min_samples_split = 8, n_estimators = 1000 | max_depth = 90, max_features = 3, min_samples_leaf = 3, min_samples_split = 8, n_estimators = 1000 |

Adaboost regressor | learning_rate = 0.1, loss = ‘square’, n_estimators = 100 | learning_rate = 1, loss = ‘linear’, n_estimators = 100 |

Multilayer perceptron | activation = ‘tanh’, alpha = 0.05, solver = ‘sgd’, hidden_layer_sizes = (100,), learning_rate = ‘adaptive’ | activation = ‘tanh’, alpha = 0.05, solver = sgd, hidden_layer_sizes = (100,), learning_rate = ‘adaptive‘ |

**Table 6.**Average result of the 10-fold cross validation method comparing the proposed uplink model versus various benchmarks including the support vector regressor, random forest regressor, Adaboost regressor, and multilayer perceptron.

IoTD-D | CUE-IoTG | |||||||
---|---|---|---|---|---|---|---|---|

MAE | RMSE | MAE | RMSE | |||||

Benchmarks | Train | Test | Train | Test | Train | Test | Train | Test |

Support vector regressor | 12.83 | 15.14 | 96.29 | 94.28 | 0.07 | 0.75 | 0.07 | 1.14 |

Random forest regressor | 2.52 | 11.63 | 35.32 | 64.84 | 0.11 | 0.83 | 0.18 | 1.16 |

Adaboost regressor | 128.06 | 129.21 | 215.24 | 216.70 | 18.13 | 18.36 | 21.69 | 21.90 |

Multilayer perceptron | 21.86 | 24.64 | 77.00 | 80.97 | 0.16 | 0.78 | 0.26 | 1.16 |

Proposed model | 9.59 | 9.84 | 66.09 | 63.43 | 0.77 | 0.77 | 1.01 | 1.06 |

**Table 7.**Average result of the 10-fold cross validation method comparing the proposed downlink model versus various benchmarks including the support vector regressor, random forest regressor, Adaboost regressor, and multilayer perceptron.

IoTG-CUE | BS-IoTD | |||||||
---|---|---|---|---|---|---|---|---|

MAE | RMSE | MAE | RMSE | |||||

Benchmarks | Train | Test | Train | Test | Train | Test | Train | Test |

Support vector regressor | 0.17 | 1.56 | 0.24 | 2.37 | 0.14 | 0.89 | 0.20 | 1.34 |

Random forest regressor | 0.26 | 1.74 | 0.39 | 2.43 | 0.16 | 0.98 | 0.24 | 1.38 |

Adaboost regressor | 40.39 | 40.75 | 49.66 | 50.16 | 21.36 | 21.69 | 25.52 | 25.83 |

Multilayer perceptron | 0.59 | 1.73 | 0.84 | 2.50 | 0.29 | 0.93 | 0.42 | 1.38 |

Proposed model | 1.64 | 1.47 | 2.16 | 2.06 | 0.94 | 0.89 | 1.25 | 1.24 |

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |

© 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Osman, R.A.; Saleh, S.N.; Saleh, Y.N.M.
A Novel Interference Avoidance Based on a Distributed Deep Learning Model for 5G-Enabled IoT. *Sensors* **2021**, *21*, 6555.
https://doi.org/10.3390/s21196555

**AMA Style**

Osman RA, Saleh SN, Saleh YNM.
A Novel Interference Avoidance Based on a Distributed Deep Learning Model for 5G-Enabled IoT. *Sensors*. 2021; 21(19):6555.
https://doi.org/10.3390/s21196555

**Chicago/Turabian Style**

Osman, Radwa Ahmed, Sherine Nagy Saleh, and Yasmine N. M. Saleh.
2021. "A Novel Interference Avoidance Based on a Distributed Deep Learning Model for 5G-Enabled IoT" *Sensors* 21, no. 19: 6555.
https://doi.org/10.3390/s21196555