A Review on Ultra Reliable and Low Latency Communications (PHY and MAC Layer Perspectives)

: Fifth Generation (5G) wireless network plays major attention in the communication scenario due to the advanced features in data transmission. 5G is designed to overcome the issues, such as data rate, reliability, throughput, latency and capacity in the wireless network domains. The 5G wireless services aim to increase the data rate through massive communication services. In this survey, well known research papers based on low latency communication at the Physical layer (PHY) and Medium Access Control layer (MAC) are reviewed based on the Ultra-Reliable Low Latency Communication (URLLC) perspective. Here, the communication techniques are categorized as, structure-based model, resource reuse-based techniques, radio resource management techniques, scheduling based approach, diversity-based techniques and network slicing model based on the purpose of data communication. The issues and challenges faced by the existing research works are reviewed and discussed in the research gaps and issues section. Moreover, the analysis carried out in this survey is based on the evaluation parameters, network layer, network parameters, system, purpose and simulation toolsets. The future work of the URLLC research is based on the issues collected from the recent research papers reviewed in this study.


Introduction
5G technology is designed using various requirements from the industrial and social sectors, where wireless technology enables the variety of diverse use cases and applications, which in turn pave the way for business models. These requirements are highly demanding in wireless technology (Colazzo et al., 2016). The 5G system will not only enhance the data rate but also meet the demands for services from industrial verticals. The network services have various needs such as network performance, like communication reliability, channel utilization, throughput and latency access (Ksentini et al., 2018). 5G network encompasses fixedmobile convergence with the connectivity speed in the multi-tenancy infrastructure. Integrating the diverse technologies with single network architecture using the network functions poses technology challenges in terms of scalability, energy efficiency and security (Siddiqui et al., 2016). However, the 5G network consists of a Heterogeneous Network (HetNet). In 5G small cells, the capacity of the network is computed to support the supermassive connection and large traffic through mmWave technology (Liang et al., 2016). 5G cellular network provides better throughput and reduced the transfer latency in the communication systems. 5G requires a better data rate in video streaming and low latency in the vehicle to vehicle transportation to ensure transportation safety (Huang et al., 2017). Network slicing is one of the key components in 5G wireless systems, where the major objective is to design the network architecture to support the Internet of Things (IoT) and massive Machine-Type Communications (mMTC), enhanced Mobile Broadband services (eMBB) and applications and URLLC communications. Besides the human-oriented smart devices, the 5G systems include vehicles, actuators and sensors to support more than 50 use case scenarios (Ksentini and Nikaein, 2017).
However, network slicing is envisioned to create the network instances, which helps to share the network infrastructure and guarantees the requirements in heterogeneous networks (Ksentini et al., 2018). URLLC is envisioned to support the automotive industry services, like industry 4.0, vehicular networks, vertical industries such as smart grid, intelligent transport and smart home environment and other services, namely tactile internet and cloud robotics. All the network services need low latency access in the communication scenario (Ksentini et al., 2018). The millimeter-Wave (mmWave) communication captured significant attention in both academia and industry, as it poses large scale antennas and wide bandwidth availability. However, mmWave backhaul is one of the important technologies used in 5G network, which is formed by combining the HetNet with mmWave communication. It is widely used to transfer heavy traffic to the internet (Liang et al., 2016). This survey is focused on reviewing various low latency 5G communication techniques at PHY, MAC and cross-layer aspects. Categorizing the considered research papers into five groups such as network slicing, scheduling, modulation, spectrum sensing and frame structure. The paper provides a common review of trends observed in the 5G research considering various aspects of communication techniques used in 5G. This survey will help the researchers to get an insight into the trends observed in the various communication techniques in 5G. Also this work focusses on clearly mentioning the approach used by other researchers considered in this study and shows how it becomes beneficial in improving the latency, bandwidth utilization and increased performance in the 5G system. This paper is organized as follows: Section 1 discusses the introduction and section 2 presents the motivation of the survey. Section 3 focusses on the related works while section 4 presents the research gaps and issues. Section 5 elaborates the analysis of various factors and finally, section 6 concludes the paper.

Motivation
The 5G

Structure-Based Communication Techniques
The research papers based on structure-based communication techniques for 5G wireless communication are discussed below: (Zhang et al., 2019) developed a joint bandwidth and latency constrained grouping algorithm to generate the network unit group for various network slice environment. It effectively provided support for latency optimization and bandwidth resource. The network unit group meets the requirements of network services to ensure bandwidth utilization. Based on the requirements of latency, the pooling cycle was shortened and reduced which leads to increase the coverage or reduce the burst transmission. The network model was validated only under fiber lengths, service ratios and network loads. Alfadhli et al. (2019) developed a latency reduction model to achieve reliable latency communication in 5G networks. It effectively enabled the data to be transmitted instantly which reduced the delay in fronthaul. Moreover, it considered some additional functions at edge nodes such as a remote unit, which in turn reduced the latency in the communication network. Correa and David (2018) modelled a custom protocol to reduce the latency and to increase the speed of communication channels in 5G networks. The utilization of channel bandwidth attained better efficiency of 89% in terms of the payload packets of 100 bytes. The communication stack of this protocol remained sufficient and simple in performing the clustering operations. However, the clock cycle waste present in the controller misused the valuable time. Raviraj et al. (2005) developed a MAC layer protocol named MOBMAC for supporting mobility in sensor networks. It used the frame size model to solve the issues of frame loss caused by the Doppler shift in mobile scenarios. The frame size used to perform the data transmission was predicted using the extended Kalman filter, which in turn reduced the latencies and increased the efficiency of the system. Zheng et al.
(2010) modeled an energy-efficient and low latency MAC protocol for data transmission in a multi-hop network. It attained better energy efficiency and lower latency. Here, the routing information was used to alter the format of the controlling frame. It utilized the carrier listening technology to involve the neighboring nodes to accomplish the process of multi-hop delivery. Zafar and Khan (2017) developed a multi cluster-based communication protocol in the network management to reduce the communication cost. It optimized the network performance by exploiting low power and low complexity in the intra-cluster and inter-cluster communication. Here, both the beaconless and beacon-enabled modes were investigated using the time division multiple access. It eliminated the access delay and collision. The summary of the above discussion has been mentioned in Table 1.

Resource Reuse-based Communication Techniques
This section elaborates on the resource reuse based 5G communication techniques gathered from various research papers. Mathur et al. (2016) introduced a crosslayer MAC and PHY layer protocol to solve the issues in bitrate limited devices. It required a wide range of data communication services and less access delay and effectively performed the operations in the same band. It allowed the IoT devices not to wait for accessing the small-sized data transmission. It reduced the control overhead, but the utilization of the underlay spectrum was considered as self-interference. Papaioannou et al. (2018) developed an optical fronthaul architecture to synergize the optical transceiver, optical multiplexer and beamforming more effectively towards the analog fronthaul in the medium transparent small cell networks. Here, the Remote Radio Heads (RRHs) offered the wireless data rates by ranging from 25 to 400 Gbps.  (2017) complexity in the intra cluster and inter cluster communication.
However, the resource allocation and small cell access were ensured through MAC protocol. It failed to monitor access control and network operations. Chung (2019) modelled a low latency and high speed passive optical network model to solve the issues in the bandwidth and latency requirements of 5G wireless services. It utilized low latency operations to increase the performance of the system. It effectively realized the flexibility for adding or replacing the new functions in the network services. Liang et al. (2016) developed a MAC protocol model for mmWave in 5G wireless network. It utilized the signaling, frame structure and backhaul or downlink or uplink resource allocation scheme at the system level in backhaul network. It efficiently connected the end-users and increased network capacity. It failed to perform backhaul routing and resource allocation in the mmWave system. Labiod et al. (2019a) developed a cross-layer model for video streaming in a wireless network. It provided better coding efficiency and low delay in the coding structure. The coding structure guaranteed the time of minimal coding in the video bitrates, which penalized the performance of the transmission. To increase the quality of the received video, the cross-layer model was utilized in vehicular communications. It offered better quality enhancements at the receiver side. However, the efficiency of video encoding was significantly poor in video prediction. Labiod et al. (2019b) developed an adaptive cross-layer model for video streaming in 5G network. The cross-layer model increased the performance of video streaming under fewer delay constraints. The transmitted video packets were assigned to the access queue in the MAC layer by considering the prediction structure. It does not perform as a classification module in the encoding configuration. Ma et al. (2017) modelled a handover authentication mechanism based on hypothesis tests in wireless communication. The fundamental parameters were designed in the physical layer to achieve the authentication process. This mechanism consumed less storage and computational resources. This framework provided a better performance in terms of security. The alarm-rate and the detection rate were significantly poor in the authentication process. Levanen et al. (2017) developed a low latency interference model to balance the amount of memory in the physical layer process of 5G network. It reduced the time duration while transmitting the frames by increasing the frequency allocation. It achieved better performance in data communication. Yang et al. (2015) developed a physical layer security model in 5G network under the frequency spectrum. It takes advantage of data confidentiality using the security paradigm in 5G. It exploited the intrinsic randomness to achieve the benefit of data security in 5G technology. It showed better performance in MIMO system. Baig et al. (2017) developed a dual physical layer transceiver in 5G to enable the pulse shaping in signal transmission. It utilized sub-layers to ensure the efficiency of 4G network devices with confinement, increased robustness and higher spectral efficiency. However, it was not effective in using wavelet filter banks. Ford et al. (2017) introduced a mmWave based low latency approach in 5G wireless cellular networks. The spectrum with highfrequency regions was effectively marked and bands with 10 GHz bandwidth were used to achieve the data communication. However, the frequency bands of mmWave support data rates and services while maintaining low latency and reliability. Levanen et al. (2014) developed a low latency interference approach in designing the 5G communication area. It provided the transmission latencies of 1 ms and increased the spectral efficiency of the system. However, the delay in the radio interface was invisible to the higher layers. Ejaz and Ibnkahla (2017) developed a multiband spectrum sensing model in the cognitive network for 5G IoT devices. The multiband model decreased the energy consumption in spectrum sensing than the single band approach. It effectively satisfies the probabilities of false alarm and detection requirements to reduce the energy by determining the number of channels in the IoT nodes. It was not suitable for mobile IoT nodes, where the channel fading alters over time. Kim et al. (2014) developed a physical layer-based MAC framework to increase the throughput of data traffic in a sensor network. optical transceiver, optical multiplexer and beamforming were ensured through MAC protocol more effectively towards the analog fronthaul in the medium transparent small cell networks Chung (2019) Developed low latency, high speed passive optical network Solved issues in bandwidth and latency Liang et al. (2016) Developed low latency, high speed passive optical network Increased network capacity. Labiod et al. (2019a) Developed a cross layer model for video streaming in Used in vehicular communication wireless network Levanen et al. (2017) Developed low latency interference model to balance the amount Achieved better performance in of memory in the physical layer process of 5G network. data communication Yang et al. (2015) Developed a PHY layer security model in 5G network Showed better performance in MIMO under the frequency spectrum. system.
Here, the acknowledgments were parallelized from sensor nodes to pull messages. The access point node knows which node has the data to forward and to allocate the channel resources by transmitting the pull message. It failed to address the strategy of group scheduling and grouping in the probe and pull mechanism. Biswash and Jayakody (2018) developed a user-centric dynamic switching and mobility management approach to increase the performance in data communication. It acts as the central node in the user mobility and switching mode during communication. However, the mode switching was based on the parameters, like delay, energy consumption and link utilization. It increased the performance of mobility management and achieved better scalability and robustness. It does not guarantee the quality of the network. The summary of the above discussion has been mentioned in Table 2.

Scheduling based Techniques
The research papers that utilized the scheduling based 5G communication techniques in URLLC are explained in this section. Ksentini et al. (2018) developed a two-level MAC scheduling model to effectively handle the downlink and uplink transmission in the network slices with various characteristics. The transmission scheduling was efficiently computed at the radio access network level. It offered more flexibility in managing the radio resources to meet the requirements of latency in communication services. However, the burden in the scheduling process at MAC layer was not resolved. Huang et al. (2017) developed a MAC scheduling model in the massive MIMO system. It enabled the processing parallelization. Scheduling in the resource blocks was executed concurrently based on the common post and pre-processing. The power-based weight is included in the scheduling to complete the link adaptation. It failed to increase throughput and transmitting power. Wang et al. (2019) developed a cross-layer resource allocation model for optimizing the rate control in the long term perspective of the wireless network. It effectively decreased the delay and increased the reliability of the network based on the relay model. It separated the crosslayer optimization problem into power allocation and rate control problem. The problem was solved using the alternating direction scheme. It failed to combine the big data with the alternating direction model to solve the complexity issues. Lähetkangas et al. (2014) modelled a 5G frame structure to initiate the data transmission or reception time, like scheduling, signaling and synchronization. With the same battery and power consumption, the frame structure reduced the energy consumption in the network design. However, utilizing the subcarrier spacing and data parts in the frame was not considered. Zhang et al. (2014) developed an energyefficient low latency model in 5G network. It utilized the top-down approach for designing the low latency and energy-efficient reliable networks. It effectively considered the air interface and user-centric scheduling to make the network as energy efficient. Pocovi et al. (2017) developed a low latency communication framework in the cellular networks. It used the low pass interference information to find the imperfection in link adaptation in sporadic traffic. It achieved a strong correlation between channel quality and user latency. However, the scheduling process and the link adaptation mechanism were not considered in the mixed traffic class. Chaudhuri et al. (2018) introduced a QoS aware MAC scheduler to adhere to the fairness and quality requirements in the network specification. Here, the multi-objective optimization was used to select the right user and their corresponding application to meet the requirements of quality. It attained better cell throughput and minimum packet drop rate in the MAC scheduler. It failed to increase the performance of video streaming users. The summary of the above discussion has been mentioned in Table 3.  (2018) adhere to the fairness and quality rate in the MAC scheduler requirements in the network specification Introduced a coding and modulation scheme Increasing the reliability for URLLC networks. for increasing the reliability for URLLC networks. Khalid et al. (2018) Developed an energy-efficient modulation Saves energy and reduces data usage by a mechanism in the 5G femtocell network to reduce thousand-fold in 5G the consumption of energy in data transmission Franchino and Buttazzo Power-aware MAC layer protocol to enhance the Effectively allocated and shared the bandwidth (2017) lifetime of the system to all the nodes in the wireless network. Hoang et al. (2014) Developed a cooperative MAC protocol for bi-It increased the performance of the network directional traffic based on the network coding through a cross-layer perspective in terms of the mode parameters such as end-to-end latency and network throughput. Sayadi et al. (2018) Introduced a radio over Ethernet protocol for Increased performance data transmission at the Ethernet frame in 5G fronthaul network

Diversity-based Techniques
This section elaborates on the diversity-based communication techniques for 5G wireless networks. Sachs et al. (2018a) introduced a coding and modulation scheme for increasing the reliability for URLLC networks. It was specified that low latency features will minimize transmission latency. However, the redundancy was provided over various transmission points and carriers with the carrier aggregation model. Khalid et al. (2018) developed an energy-efficient modulation mechanism in the 5G femtocell network to reduce the consumption of energy in data transmission. The adaptive modulation saves energy and reduces data usage by a thousand-fold in 5G. Franchino and Buttazzo (2017) introduced a power-aware MAC layer protocol to enhance the lifetime of the system by bounding the delay of real-time messages. It effectively allocated and shared the bandwidth to all the nodes in the wireless network. The protocol performed multihop communication in various network topologies. However, it failed to perform the delay analysis in case of downstream flow. Hoang et al. (2014) developed a cooperative MAC protocol for bi-directional traffic based on the network coding mode. It increased the performance of the network through a cross-layer perspective in terms of the parameters such as end-toend latency and network throughput. The performance was validated using Monte-Carlo simulation. It failed to integrate the network layer with the cross-layer in a wireless network. Sayadi et al. (2018) introduced a radio over Ethernet protocol for data transmission at the Ethernet frame in 5G fronthaul network. Here, the mapper present at the transmitter side received the data and forward it to the Ethernet frame. The mapper situated at the destination end received the Ethernet frame and extract the data. However, running the mapper in the embedded system failed to reduce the delay in data transmission. The summary of the above discussion has been mentioned in Table 4.

Radio Resource Management Methods
This section elaborates on the radio resource management techniques in 5G wireless network collected from the existing research papers. Roth-Mandutz et al. (2017) developed a self-optimized model at the network management level for reducing the capacity impact in ultra-reliable communication services. It effectively selected the appropriate parameters and identified the legacy use cases. It enhanced the performance by meeting the diverse requirements. However, it failed to enhance the parameter optimization at the user-centric and service network. Colazzo et al. (2016) developed a Novel Radio Multiservice Adaptive network model (NORMA) in the 5G network to achieve low latency wireless communication. Here, the network slice effectively allocated the network functions to be required in the radio interface. Based on the lean protocol stack, the packets were able to pass over the stack layers more efficiently. Dighriri et al. (2016) developed a data traffic model in 5G network to achieve machine to machine communication. Based on the measure and classification of data traffic, the quality of service in the network slicing was effectively satisfied in smart systems. It efficiently utilized the radio resources in the relay node of the physical resource block. The fairness of service and the data traffic explosion was not resolved. Rony et al. (2018) introduced a radio access network based architecture in 5G network. The functionality split at the physical layer effectively tackled the challenges associated with the radio access network. It supports the centralization of the upper layers to achieve a high data rate. It was not effective in traffic load conditions. Shreevastav and Carbajo (2016) developed an adaptive radio link control protocol to increase the throughput of network modalities in a cellular system. It reduced the processing power and increased the system latency by enabling the un-acknowledgment mode. However, the radio condition activates the acknowledgment mode to increase data reliability. It was not suitable in 5G split architectural environment. The summary of the above discussion has been mentioned in Table 5.

Network Slicing Techniques
Some of the network slicing methods collected from the existing research papers are represented in this section.  introduced a network slicing model for 5G communication systems. This model was based on the end to end horizontal and vertical slicing, where horizontal slicing increases the user experience and system capacity whereas vertical slicing enables the vertical services. However, the communication and the computation resources were horizontally sliced to generate the virtual computation environment to solve the network scaling problems. However, the granularity and the ways of doing network slicing were not primarily focused. Ksentini and Nikaein (2017) developed a network slicing model in the radio access network. This model relies on the principle of a dedicated core network to separate the traffic and to share the resources between the slices. It enabled the dynamicity and flexibility to enforce the network slicing in the radio access network. However, the two-level scheduling processes for the stability, performance and scalability were not addressed. Rost et al. (2017) developed a network slicing model to design the 5G mobile networks. It effectively analyzed the realization options in the 5G network based on the network slices. It offered revenue to the network operator and achieved mapping with the shared and the dedicated slices. Siddiqui et al. (2016) introduced virtualized intelligent and hierarchical network architecture. It explored the hierarchical routing in the network caching to achieve the requirements of low latency. It effectively solved the issues of multi-tenancy and security in the network slicing. It failed to reduce the content-based access latency in fixed and mobile networks. Herzog et al. (2016) developed a network slicing paradigm to enable agile, scalable and highly flexible 5G compatible network architecture.MAC layer can support the multi-radio access operations to maximize the traffic volume, which effectively enhanced the quality of different users in the network system. However, partitioning and managing the resources across the heterogeneous networks was not achieved. Chen et al. (2019) developed a network slicing based resource management model to modify the control model of network architecture. For event and application types, the network resources offered suitable prioritization and bandwidth and ensure the reliability of data transmission. It failed to consider the stability of the system even under abnormal events. The summary of the above discussion has been mentioned in Table 6.  Colazzo et al. (2016) Developed a novel radio multiservice adaptive Low latency wireless communication network model (NORMA) in the 5G network Dighriri et al. (2016) Developed a data traffic model in 5G network It efficiently utilized the radio resources in the relay node of the physical resource block Rony et al. (2018) Introduced a radio access network based Tackled the challenges associated with the radio architecture in 5G network access network Shreevastav and Carbajo Developed an adaptive radio link control Increase the throughput of network modalities in (2016) protocol a cellular system   Introduced network slicing model for 5G Increases the user experience and system capacity communication systems. Ksentini and Nikaein Developed network slicing model in the It enabled the dynamicity and flexibility to enforce the (2017) radio access network. network slicing in the radio access network Rost et al. (2017) Developed a network slicing model It effectively analysed the realization options in the 5G network based on the network slices Siddiqui et al. (2016) Introduced virtualized intelligent and Achieved the requirements of low latency hierarchical network architecture. Herzog et al. (2016) Developed a network slicing paradigm to Enhanced the quality of different users in the network enable agile, scalable and highly flexible system 5G compatible network architecture Chen et al. (2019) Developed a network slicing based resource Offered suitable prioritization and bandwidth and ensure management model to modify the control the reliability of data transmission model of network architecture Other 5G Communication Techniques network. It defines the mathematical formulation at the network edge or fog nodes in the placement of service infrastructure. It effectively solved the optimization issues based on the position of traffic generators. The placement parameters like location cost and candidate sites were not considered. Moreira et al. (2018) developed a cross-layer authentication protocol to solve the issues in weak security and latency. The radio trusted database aims to increase the authentication of radio devices present in the wireless network. However, the entire coverage area is partitioned into various trusted zones, where each zone consists of numerous cells to initiate the authentication request. It reduced the computational complexity and minimized the recurrence of authentication. It failed to classify the profiles of HetNet to design efficient and reliable trusted zones. Al-Raweshidy et al. (2017) introduced a new intelligent model for optimally selecting the parameter sets in MAC layer. However, the optimized parameters considerably attained higher throughput. Here, the network parameters were neither validated nor simulated in the real-time scenario.

Research Gaps and Issues
The research gaps and issues faced by the considered research papers are elaborated in this survey. The summary of the research gaps and issues has been described in Table 7.  Ksentini et al. (2018), the utilization of the underlay spectrum was self-interference and the based techniques access control and network operations were not effectively monitored in (Papaioannou et al., 2018).  It failed to perform backhaul routing and resource allocation in the mmWave system in (Liang et al., 2016).  The efficiency of video encoding was significantly poor in video prediction in (Labiod et al., 2019a). However, the classification module was not performed in the encoding configuration. The Alarm rate and the detection rate were significantly poor in the authentication process (Ma et al., 2017) and the resource model in (Yang et al., 2015) was not effective in using the wavelet filter banks.  The channel fading altering with time was not suitable for mobile IoT nodes (Ejaz and Ibnkahla, 2017). It failed to address the strategy of group scheduling and grouping in the probe and pull mechanism (Kim et al., 2014).  The quality of the network was not guaranteed in (Biswash and Jayakody, 2018).

Scheduling based
 The burden in the scheduling process at MAC layer was not resolved in (Zhang et al., 2014) techniques and the throughput and transmitting power is reduced in (Huang et al., 2017).  It failed to combine the big data with the alternating direction model to solve the complexity issues according to (Wang et al., 2019).  Lähetkangas et al. (2014), the subcarrier spacing and data parts are failed to use in the frame.  Siddiqui et al. (2016), the performance was poor for video streaming users.

Analysis Based on the Network
The analysis carried out based on the category of the network is discussed in this section. Most of the papers discussed here consider 5G wireless networks. Four research papers used 5G mobile networks whereas the radio access network is used by one research paper. However, 5G cellular networks, 5G Hetnet, Flying Ad hoc Networks (FANET) and Software Defined Network (SDN) are used by a single research paper. The vehicular ad hoc networks (VANET) are used by two research papers and so on. The analysis based on the network is represented in Fig. 2.

Analysis based on Purpose in 5G Communication
The analysis carried out based on the purpose in 5G network communication is discussed in this section. Here, network communication is widely used in eight research papers and the purpose of network slicing is used in four research works. Authentication, video streaming, beamforming and scheduling are preferred in three research papers. From Fig. 3, it is depicted that network communication is widely preferred in most of the research works.

Analysis based on Simulation Tool Used
The analysis carried out based on the simulation tool used is discussed in this section. Here, the MATLAB and the Network Simulator 2 (NS-2) is the widely used simulation tools. Moreover, the vehicle simulator, OPNET network simulator and Automated Validation of Internet Security Protocols and Applications (AVISPA) tools are used in two research papers and the NS 3 simulator is used in three research paper. From Fig. 4, it is clearly shown that MATLAB is a widely preferred tool in 5G communication URLLC research. However, adding NS 2 and NS 3 outperforms MATLAB.

Analysis based on the System
In this section, the analysis performed based on the system is discussed. Here, three research papers used the mmWave and MIMO systems. IoT and the wireless systems are used in two papers. Embedded systems, multi-user cellular systems and the mobility management system are used in single research work. Figure 5 portrays the analysis based on the system.

Analysis based on Communication Techniques
This section describes the analysis performed based on the categorization of existing methods. The resource reuse-based techniques are used in fifteen research papers and scheduling based techniques are used in seven works. Network slicing model, structure-based and other techniques are used in six research works. From Fig. 6, it is depicted that the resource reuse method is widely preferred.

Analysis Based on Network Parameter
This section elaborates the analysis carried out based on the network parameter in 5G wireless communication.
Here, the network parameters, like latency and reliability are focused and the existing research papers are categorized based on the network parameters. Here, latency is used in 33 research papers, while reliability is used in two research works. Figure 7 depicts the analysis based on network parameter.

Analysis Based on Network Layer
This section elaborates the analysis based on the network layer. Here, the PHY layer is used in 17 research papers, MAC layer is used in 12 research works and cross layer is used in nine research papers. It is clearly pictured that PHY layer is used in most of the research papers. Figure 8 portrays the analysis based on network layer.

Conclusion
The review in overall shows a general trend of the current research happenings. All the survey conclusions are based on the considered research papers. We found that 68% of research work was focused on 5G networks whereas the second highest of 9% research work was carried out in the 5G mobile wireless networks. Considering the purpose of communication in 5G, we found that the maximum research papers were focused on network communication. Most widely used simulation tools within these research papers were the MATLAB and the network simulator 2 (NS-2). Considering the analysis based on system, mmWave and MIMO systems were found to be considered maximum in this survey. Considering the analysis based on communication technique, we found that 15 research papers considered resource reuse based technique whereas scheduling techniques were used second most.