5g Cloud-Radio Access Network Intra-Phy Functional Split Analysis

375

5g Cloud-Radio Access Network Intra-Phy

Functional Split Analysis

Ebude Carine Awasume, Stephen Musyoski, Vitalice Kalecha Oduol

Abstract: Cloud-RAN (CRAN) is considered as one of the proposed solutions for 5G networks, enabling new wireless technologies such as coordinated multi-point (CoMP), advanced intercell interference coordination (ICIC), massive multiple-input multiple-output (MIMO), etc. However, the main challenge in the implementation of the CRAN architecture is that it requires very high capacity and low latency fronthaul (FH) links to carry raw I/Q samples between remote radio heads (RRH) and the baseband units (BBUs). These requirements can be reduced by adjusting the functional split of baseband processing between BBUs and RRHs. The process of moving some of the functions from a traditional cloud RAN’s BBU into the RRH reduces the fronthaul rates. Therefore, in this paper, we propose an intra-PHY functional split that will drastically reduce the fronthaul bandwidth by more than ten times and still presents most of the advantages of centralized processing which is a principal objective of the C-RAN technology. As future work, an appropriate fronthaul network will be established with a dynamic bandwidth allocation algorithm to reduce the fronthaul latency.

Index Terms: Cloud Radio Access Network, Common Public Radio Interface, Data rate, Fifth generation, Fronthaul, Functional split, latency

——————————  ——————————

1.

INTRODUCTION

Mobile traffic demand has increased drastically in recent days and with such growth, studies and research are being done to cope with the challenges of the communication systems industry[1]. This increase in data rate requirement puts urgency on network operator side to increase data rate without compromising the cost and quality of service[2]. Cloud-Radio Access Networks Architecture (C-RAN) is one of the proposed solutions to address the increased demand, and is a potential candidate for future 5G networks[3]. In Long Term Evolution (LTE) architecture, eNodeB (eNB) contains two main parts: Baseband Unit (BBU) and Remote Radio Head (RRH). RRHs transform the baseband signals from BBUs to radio frequency and then forward it to User Equipments (UEs) by the antennas in the downlink. The process is adverse in the uplink. Meanwhile, BBUs deal with baseband signal processing, RRH is connected to BBU through optical fibers[4] Contrarily to conventional architecture, the C-RAN architecture divides the entire function of the traditional base station into two elements: the Base band Unit (BBU), consisting in an intelligent element to perform baseband tasks functionalities, and the Remote Radio Head (RRH), that consists of a passive antenna element to provide access for serviced Users Equipment (UEs)[5]. The BBU pool is connected to hundreds or thousands of RRHs via the fronthaul network. Fig. 1 illustrates a network overview of C-RAN architecture.

Fig. 1. Network overview of C-RAN architecture illustrating BBU-pool, RRH and the fronthaul network[6].

interference coordination (ICIC), massive multiple-input multiple-output (MIMO), etc.[7] because of the centralized processing of the radio signals. It also has the ability to establish significantly lower costs[8] and greener communication[9]. However, 5G C-RAN architecture also presents some challenges as any other network technology. In particular, development of a reliable fronthaul network with required capacity and delay for a large number of cells in a cost and energy efficient manner is one of the major challenges[10]. Optical network has been considered the best solution for a 5G fronthaul network amongst several wired and wireless possible fronthaul technologies. This is because of its properties of low latency and high capacity. This fiber-wireless (Fi-Wi) convergence has to be designed carefully as it an important factor which contributes to end-to end delay, cost and capacity of the C-RAN network.

In order to relax the excessive FH capacity constraint, changing the current functional split architecture between RRH and BBU has been considered as one of the promising solutions to overcome such high bandwidth and tight latency requirements[11]. The rest of this paper is organized as follows: in section 2, we presented the implementation challenges of 5G C-RAN. Section 3 presents the requirements of the traditional RRH-BBU functional split. Section 4 analysis our proposed Intra-PHY functional split. Section 5 then provides the simulation results and analysis while section 6 ————————————————

 Ebude Carine is currently pursuing PhD program in electrical engineering (Telecom option) in the Pan African University Institute of Basic Sciences, Technology and Innovation (PAUISTI), Nairobi, Kenya. E-mail: [email protected]

 Stephen Musyoski (PhD) B.Sc, M.Sc is an Associate Professor in the Department of Telecommunications and Information Engineering, Technical University of Kenya, Kenya

376 concludes the paper.

2

5G

C-RAN

I

MPLEMENTATION

C

HALLENGES

Any 5G C-RAN architecture consists of three main components which are shown in Fig. 1. RRHs are deployed closer to the user equipment and are connected to the centrally controlled BBU pool over the fronthaul network. Then, the BBU pool is connected through backhaul to the core. The C-RAN architecture has been proven to be efficient in reducing CAPEX/OPEX, power consumption and duration of construction cycle[12]. In C-RAN, the transport network affects the capacity, latency, and level of intelligence of the network. Therefore, development of architectures, technologies, interfaces, and networks for 5G fronthaul has gained significant attention from both academia and industry in last few years. Different[9]. Standardization bodies such as 3GPP and IEEE Next- Generation Fronthaul Interface (NGFI) working group have been actively investigating on different functional split architectures. A functional split is what determines the number of functions left locally at the RRH, and the number of functions centralized at the BBU. In NGFI[13], various functional splits are being defined to provide different trade-offs among RRH complexity, system performance, fronthaul capacity and fronthaul bandwidth efficiency. In order to relax the excessive FH capacity constraint, changing the current functional split architecture between RRH and BBU has been considered as one of the promising solutions to overcome such high bandwidth and tight latency requirements[11]. Normally higher level split interfaces, which distribute more functionality toward RRHs, would increase bandwidth efficiency. However, that would also kill the advantages of centralized baseband radio processing. A carefully engineered fronthaul interface is required to maximize the benefit of centralization while constraining the fronthaul networking requirement. 3GPP has started studying different functional splits between Central Unit CU and Distributed Unit DU. For the initial phase, 3GPP has taken LTE protocol stack as a basis for the discussion, until RAN2 defines and freezes the protocol stack for 5G New Radio (NR). About 8 possible options including several sub-options have been proposed as shown in Fig. 2. The different functional split options are indicated with the red lines within Fig. 2. The functions found below the red lines shall be implemented in the DU while those found above shall be in the CU. The more functions located in the DU, the more processing has already been done before data is transmitted on the fronthaul network, and the lower bitrate on the fronthaul network[6].

Fig 2: The LTE protocol stack with layers and sublayers, including the numbered functional split options proposed by

3GPP[6]

3

THE

TRADITIONAL

RRH-BBU

SPLIT

REQUIREMENTS

This is the traditional functional split also known as the option 8(PHY-RF split) functional split. This option allows to separate the RF and the PHY layer resulting in a very simple RRH. This split permit centralization of processes at all protocol layer levels, resulting in very tight coordination of the RAN. This allows efficient support of functions such as CoMP, MIMO, load balancing, mobility[14]. This option may be more robust over non-ideal transmission conditions and during mobility, because the ARQ is centralized in the CU[15].

The advantages of this split include:

1. High levels of centralization and coordination across the whole protocol stack, which may enable a more efficient resource management and radio performance

2. Separation between RF and PHY enables to isolate the RF components from updates to PHY, which may improve RF/PHY scalability

3. Separation of RF and PHY will allow the reuse of the RF components to serve PHY layers of different radio access technologies (e.g. GSM, 3G, LTE)

4. Separation of RF and PHY allows pooling of PHY resources, which may enable a more cost efficient dimensioning of the PHY layer

5. Separation of RF and PHY allows operators to share RF components, which may reduce system and site costs[15]

The radio signals here are being transmitted in the fronthaul network using mostly the Common Public Radio Interface (CPRI) protocol which encapsulates raw IQ samples over fiber links. The transmission of IQ samples requires a very large bandwidth and this is the main disadvantage of this split. For example, a current LTE base station which supports 150 Mbps of downlink bandwidth in the access network will require more than 2 Gbps of optical bandwidth to send its IQ samples over the CPRI interface[9]. What follows is an analysis of the CPRI being used in a 5G C-RAN.

3.1 Data Rate

377 614.4Mbit/s up to 10.137Gbit/s depending on radio access

technology (RAT), carrier bandwidth and Multiple Input Multiple Output (MIMO) implementation[16]. One of the major evolutions in 5G is the use of massive MIMO techniques to increase radio bandwidth. The CPRI data-rate for multi-sector and multi-antenna configurations is expressed as[17]

𝐷𝑎𝑡𝑎 𝑟𝑎𝑡𝑒 = 𝑆 × 𝑁 × 𝑓 × 𝑁 × 2 × 𝐶 × 𝐶 (1) where 𝑆 denotes the number of sector, 𝑁 represents the number of antennas per sector, 𝑓 denotes the sampling rate used for digitization (sample/s/carrier), 𝑁 is the sample width (bits/sample), (2) denotes a multiplication factor of two introduced to account for the (IQ) data, 𝐶 is the factor of CPRI control word (CW) and 𝐶 is a coding factor (either 10/8 for 8B/10B code or 66/64 for 64B/66B code). A typical example of values that can be used for data rate calculations for an LTE and a 5G network are sho

wn in table 1.

TABLE1

PARAMETERS FOR CPRIBANDWIDTH CALCULATION

Parameters LTE values 5G values

𝑆 3 3

𝑁 1 8

𝑓 30 30

𝑁 30.72 (for 20 MHz) 30.72×5(for 100MHz)

𝐶 16/15 16/15

𝐶 10/8 10/8

From the above table, in order to transport data with an 8X8 MIMO antenna from 3 sectors, 294.9 Gbps of optical fiber link is required for a CPRI-based fronthaul network. The fixed (i.e does not vary with real time traffic load) bandwidth requirement for CPRI fronthaul technology constitutes its major disadvantage. For example, even RRHs do not have user traffic all the time, an optical link that has a constant bandwidth needs to be dedicated for each RRH. The fronthaul capacity requirement of CPRI links can be reduced using different compression techniques[18][19]. Even with these compression techniques, a significant amount of fronthaul traffic will still be required. Therefore, an appropriate fronthaul transport technology like the use of PONs need to be examined.

3.2 Latency

In addition to the fronthaul data rate requirement, there isthe strict requirement for latency in the 5G fronthaul network. Some of the 5G network applications are highly delay sensitive. C-RAN applications such as coordinated multipoint (CoMP) and virtual BBU migration are time-sensitive, low latency should be guaranteed for these services. The acceptable round-trip delay over the fronthaul network should be ≤500μs for URLLC and ≤4ms for eMBB, including the fiber propagation delay and equipment delay. For these to be achieved, different fronthaul technologies and architectures have to be considered. The 5G networks have an even more strict delay budget for fronthaul. For example, the BBU round trip processing delay and RF processing delay can be as large as 2 ms. Therefore, with a safe margin the fronthaul network delay can only increase by a few hundreds of microseconds including the propagation delay, round trip CPRI processing delay, and the other fronthaul equipment processing delays[9]. All these will result in a constraint on the maximum delay period of each fiber- based fronthaul architecture. If the CPRI processing delay is supposed to be ≤100µs, then the CPRI-based fronthaul architecture, RRH can be placed within 20km

from a BBU considering the delay in light propagating in an optical fiber. From these, we can conclude that if delay of fronthaul is increased by few hundreds microseconds, the strict latency requirement will not be achieved. Therefore, fronthaul networks will always have strict latency requirement similar to that of the CPRI link no matter the functional split considered.

4

PROPOSED

INTRA

PHY

FUNCTIONAL

SPLIT

The proposed functional split is an intra-physical functional split. This decision was made because this will drastically reduce the fronthaul data rate and still presents most of the advantages of centralized processing which is a principal objective of the C-RAN technology. The proposed functional split is shown in Fig. 3. In this split, all of the RLC and MAC layer protocols, and some of the physical layer functions like the wireless channel coding are implemented in the centralized BBU pool. However, all the other PHY layer functions below wireless channel coding are placed in the RRH. This split gives a significant lower bitrate on the fronthaul link. This is because the signals sent have already been modulated which means that several bits (depending on the modulation order) are coded with a symbol before transmission on the fronthaul link. As a result, with this functional split, code words are being sent through the fronthaul link. Other advantages of this functional split include: 1. The close relation between the FEC and MAC layer is

kept.

2. Reduced cost for fronthaul deployment because of reduction in fronthaul transmission bandwidth and resources.

3. Fronthaul link load is not constant like in CPRI but varies with the different cell load.

4. Centralized scheduling is still possible.

5. Joint transmission and joint reception are also possible.

378

Fig. 3: The proposed Intra-PHY functional split

The maximum downlink bandwidth requirement for our intra PHY split can be calculated as follows[9],

𝐵 = 𝑀 × 𝑁 × 𝑁 × 𝑁 × 𝑁 × 𝑁 \ 𝑇𝑇𝐼 (2)

where TTI is the transmission time interval which is the duration of radio signal transmission, 𝑀 is the highest modulation order of the radio signal, 𝑁 is the number of symbol within a TTI, 𝑁 is the number of subcarriers in a resource block, 𝑁 is the number of resource blocks, and 𝑁 is the number of MIMO streams. Typical 5G network implementation values for, TTI, M, Ns, Nsc, Nrb, and Nmimo can be chosen as 1ms, 8 corresponds to 64 QAM, 12, 12, 500, and 8, respectively. Considering the above data, the maximum fronthaul bandwidth will be 4.6Gbps which is a drastic decrease compared to its CPRI counterpart.

5

S

IMULATION

R

ESULTS AND

A

NALYSIS

.

In the simulation to evaluate the MFH optical transmission bandwidth, it was assumed that many users were randomly distributed in a cell area and each user selected its own modulation scheme according to its received signal to noise ratio (SNR). The different simulation parameters shown in Table 2 below[20].

TABLE2 SIMULATION PARAMETERS

Parameter Values

System bandwidth 100 MHz (20 MHz x 5)

Number of MIMO streams _{4 (downlink), 2 (uplink)}

Sampling frequency 30.72 MHz

Number of antennas in RRH 8

Number of Resource Blocks 500

Number of subcarriers per

resource block 12

Number of allocated

subcarriers per user 60

Number of quantization bits for

IQ sample 15

Number of quantization bits for

likelihood values 2

Fig. 4 shows a plot of the mobile FH bandwidth requirement against number of users. This result further confirms that the bandwidth requirement for CPRI is independent of the number users. The bandwidth required for Intra-PHY is dependent on the number of users at each time. It is also very clear that CPRI needs more than ten times the bandwidth required for our proposed intra-PHY split. This architecture can also obtain statistical multiplexing gain since the FH transmission bandwidth varies according to mobile traffic.

Fig. 4: Comparison of FH bandwidth between CPRI and Intra-PHY

With respect to the latency requirement, there is an increase by a few micro seconds in the latency of our Intra-PHY split compared to CPRI because of the additional symbol level processing function added to the RRH. However, the additional delay is small due to the fact that the modulation delay, RF processing delay and propagation delay should be less than 5 µs, which is the cyclic prefix of an OFDM symbol[6].

6

CONCLUSION

379 an increase by a few micro seconds in the FH latency. As

future work, an appropriate fronthaul network will be established with a dynamic bandwidth allocation algorithm to reduce this fronthaul latency for the requirement to be met.

ACKNOWLEDGMENT

This work was supported by the Pan African University Institute of Basic Science, Technology and Innovation PAUISTI, Kenya.

REFERENCES

[1] H. Alsaif, ―Extreme Wide Band MIMO Antenna System for Fifth Generation Wireless Systems,‖ Eng. Technol. Appl. Sci. Res., vol. 10, no. 2, pp. 5492–5495, 2020. [2] A. Shabbir, H. R. Khan, S. A. Ali, and S. Rizvi, ―Design

and Performance Analysis of Multi-tier Heterogeneous Network through Coverage, Throughput and Energy Efficiency,‖ Eng. Technol. Appl. Sci. Res., vol. 7, no. 6, pp. 2345–2350, 2017.

[3] C. Chang, R. Schiavi, N. Nikaein, T. Spyropoulos, and C. Bonnet, ―Impact of Packetization and Functional Split on C-RAN Fronthaul Performance,‖ in 2016 IEEE International Conference on Communications (ICC), 2016, pp. 1–7.

[4] J. Duan, X. Lagrange, and F. Guilloud, ―Performance Analysis of Several Functional Splits in C-RAN,‖ in 2016 IEEE 83rd Vehicular Technology Conference (VTC Spring), 2016, pp. 1–5.

[5] X. Zhang, M. Peng, Z. Wang, and S. Mai, ―Performance Analysis and Evaluation of Cloud Radio Access Networks,‖ in 2018 International Conference on Computing, Networking and Communications (ICNC), 2018, pp. 68–72, doi: 10.1109/ICCNC.2018.8390348.

[6] L. M. P. Larsen, A. Checko, and H. L. Christiansen, ―A Survey of the Functional Splits Proposed for 5G Mobile Crosshaul Networks,‖ IEEE Commun. Surv. Tutorials,

pp. 146–172, 2018, doi:

10.1109/COMST.2018.2868805.

[7] S. Zhou, X. Liu, F. Effenberger, and J. Chao, ―Low-Latency High-Efficiency Mobile Fronthaul With TDM-PON ( Mobile-TDM-PON ),‖ J. Opt. Commun. Netw., vol. 10, no. 1, pp. 20–26, 2018.

[8] K. Miyamoto, T. Shimizu, J. Terada, and A. Otaka, ―Low Latency Symbol Level Transmission Scheme for Mobile Fronthaul with Intra PHY Split RAN Architecture,‖ in 2017 European Conference on Optical Communication (ECOC), 2017, pp. 1–3.

[9] C. Ranaweera, E. Wong, A. Nirmalathas, C. Jayasundara, and C. Lim, ―5G C-RAN Architecture : A Comparison of Multiple Optical Fronthaul Networks,‖ in 2017 International Conference on Optical Network Design and Modeling (ONDM), 2017, pp. 1–6.

[10] H. Wang, M. M. A. Hossain, and C. Cavdar, ―Cloud RAN Architectures with Optical and mm-Wave Transport Technologies,‖ in 2017 19th International Conference on Transparent Optical Networks (ICTON), 2017, pp. 1–4, doi: 10.1109/ICTON.2017.8025007. [11] H. Touati, H. Castel-Taleb, B. Jouaber, and S.

Akbarzadeh, ―Split analysis and fronthaul dimensioning in 5G C-RAN to guarantee ultra low latency,‖ in 2020 IEEE 17th Annual Consumer Communications & Networking Conference (CCNC), 2020, pp. 1–4, doi:

10.1109/ccnc46108.2020.9045398.

[12] Next Generation Mobile Network Alliance, ―5G White Paper,‖ 2015. [Online]. Available: https://www.ngmn.org/work-programme/5g-white-paper.html.

[13] China Mobile Research Institute, ―White Paper of Next Generation Fronthaul Interface,‖ 2015. [Online]. Available: https://docplayer.net/46899964-White-paper-of-next-generation-fronthaul-interface.html.

[14] S. R. Kumer, ―Functional Split Options in C-RAN

Architecture,‖ 2018. doi:

10.13140/RG.2.2.31060.04480.

[15] 3GPP, ―3GPP TR 38.801: Study on new radio access technology: Radio access architecture and interfaces,‖ 2017.

[16] A. Pizzinat, P. Chanclou, F. Saliou, and T. Diallo, ―Things You Should Know About Fronthaul,‖ J. Light. Technol., vol. 33, no. 5, pp. 1077–1083, 2015, doi: 10.1109/JLT.2014.2382872.

[17] I. A. Alimi, A. L. Teixeira, and P. P. Monteiro, ―Toward an Efficient C-RAN Optical Fronthaul for the Future Networks: A Tutorial on Technologies, Requirements, Challenges, and Solutions,‖ IEEE Commun. Surv. Tutorials, vol. 20, no. 1, pp. 708–769, 2018, doi: 10.1109/COMST.2017.2773462.

[18] H. Lehrmann, M. Stübert, A. Checko, H. L. Christiansen, and Y. Yan, ―Cloud RAN for Mobile Networks - a Technology Overview,‖ I E E E Commun. Surv. Tutorials Link, 2014, doi: 10.1109/COMST.2014.2355255.

[19] B. Guo, W. Cao, A. Tao, and D. Samardzija, ―LTE / LTE-A Signal Compression on the CPRI Interface,‖ Bell Labs Tech. J., vol. 18, no. 2, pp. 117–133, 2013, doi: 10.1002/bltj.