Dr C.S. Rao, Chairman and Co-Founder, QuadGen Wireless Solutions
Telco operator network (NW) deployment trends are undergoing a fundamental shift, and operators are digitally transforming their NW to meet the aspirations of digital era services and applications. The drivers towards the evolution of telco cloud are rapid time to market (TTM) for new technology adoption, new NW architectural migration and new service pack offering to subscribers is with absolute focus on reduction in the total cost of ownership (TCO). Telcos have a need to adapt to the co-existence era of global OTT players. Thus, telcos are facing additional challenges from web and content providers in the era of convergence of the Telecom, Media and Technology (TMT) industries, which are at the centre of a Digital Networked Economy (DNE). Telco Cloud is an evolution triggered by the transformation of traditional Radio Access Network (RAN), IP-OTN (IP Optical Transport NW) and Core NW Architectural Layers. Standardisation, softwarisation, virtualisation and automation trends are driving the cloudification era of the telco NW. Telcos are adopting on-premises, or public cloud, or hybrid cloud adoption models driven by the service level agreement (SLA), security and maturity of cloud operator services and cloud NW services capability. A key component of telco cloud is the IP Multimedia System (IMS), which provides convergence of voice, data, video, multimedia messaging, etc. with its integral OSS/BSS/CRM applications. 3GPP standards are also driving the new era of NW architectural support through oRAN, dRAN, vRAN and cRAN with corresponding SDN-NFV technologies. This article discusses the digital transformation trends with the deployment criteria for a global class digital telco cloud evolution.
Digital transformation in telco NW
Digital stands for fast-paced change with rapid adoption of state-of-the-art technologies like 5G in mid-band spectrum and mmWave bands, Wi-Fi-6, E-band, Terabit IP-OTN and SDM-NFV-based core network. Transformation stands for change through NW infra evolution an innovation in NW operation through reinventing tools for automation. NW operational launch digitisation through the next-gen suite of engineering services involves real-time customisation of 3As, i.e., Automation centric, Analytics based, Agile response for Service Assurance focus. Digital transformation involves five technology pillars of the digital decade, i.e., 5G with coexisting Wi-Fi-6/7, SDN-NFV-based oRAN, dRAN, vRAN and cRAN, cloud native architecture, AI-enabled NW operations and optimisation with E-band wireless backhaul.
AI adoption drives radio resource management (RRM) on UL/DL, CSI (Channel State Information) monitoring to proactively measure the properties of the radio channel to ensure proper modulation, code rate, beam shaping, tracking and acquisition for best signal quality and data throughput. Traditional integrated gNB in the RAN layer is getting deployed as a dRAN with BBU (Base Band Unit) functions like control unit (CU) and data unit (DU) moved to the multi-edge cloud (MEC) or central cloud (CC) platform. This BBU including CU and DU functions has become virtual network functions (VNF) in the SDN era and hosted on MEC or CC. Thus, the RAN layer has migrated to the cloud era for faster adoption. 
Cellular radio access traffic from the BTS has reached a huge capacity at 2 Gbps to 6 Gbps and this necessitates point to point (PTP) E-band backhaul links in addition to 10 Gbps IP-OTN Tx links. Digital transformation in four layers, i.e., RAN, Tx, Core and OSS layers, has taken place including the 5th dimension in the business services transformation and customer experience (CX) domains.
Telcom service providers (TSPs) have reinvented themselves into application service provider (ASPs), CATV providers, digital store providers (DSPs), content service providers (CSPs), security service providers (SSPs) and digital private network (DPN) service providers. All this has become possible because of cloud native architecture adoption with regard to SaaS, WaaS, CaaS, NaaS, STaaS, LaaS, etc., leading to a XaaS era driven by AI -ML adoption.
Business services to varied customers like m-everything require superfast automated zero touch NW element configuration, fault management and E2E integration demands to enable the operator to evolve as a digital telco by embracing the latest NG OSS capabilities including self-optimised NW. A unique digital operator architecture is in evolution with a cloud, SDN/NFV and AI/ML-led path on massive NEs of high capacity in every NW layer, thus looking for a predictive NES with informatics and analytics era-based real-time automated NW Mtce with much less human intervention of massive scale of NW elements.
It is imperative to know the scale of transformation in technological functions and features. Mobile broadband (MBB) has migrated to enhanced MBB (eMBB) at a scale of 100x to 1,000x. Machine type communications (MTC) has migrated to massive MTC (mMTC) at a scale of 10,000x. Reliable communications (RC) has migrated to Ultra RC (URC) at a scale of 1,000x. Latency has migrated to low latency at a 1/50 scale. Smartphone (SP) devices have migrated to mobile computing platforms with 3x CPU cores including GPU and devise memory scaled to 5x levels. The fibre Tx domain has migrated to Packet Over OTN (IP-OTN) at a scale of 1,000x. QoS has migrated to QoE at a granular scale of SLA. This scale of NW infra evolution drove the compelling need for the adoption of cloud native architecture with AI-ML in NW operations from OSS/BSS to CRM. 
The SP-led era is giving way to a whole host of services on embedded things. Digital transformation in UE with built-in sensor and app-driven functions support multiple services like health and fitness, smart offices, G2C, EV apps, etc.
RAN transformation involved the RF channel width increasing at a scale of 40x from 10 MHz to 400 MHz. The antenna order changed from 2×2 to 32×32 with beam forming and antenna arrays being adopted as active antennas. The number of sectors on the BTS has scaled up from 3 to 12.
gNB has just become a radio unit (RU), a lightweight and low power consumption with All In One (AIO), with a built-in active phased array antenna unit (AU), capable of being mounted on an 8 metre electric pole serving a residential cluster.
Cellular Backhaul Tx Network Architecture has changed from one haul to three hauls, i.e., front-haul/mid-haul and backhaul. The speeds of transmission layers have scaled up by 40x, i.e., 1G to 40G, 40G to 400G, 400 Gbps to 16 Tbps.
Digital transformation in the 5G core NW involves migration from physical NW elements and NW interfaces giving way to virtual functions such as AMF, SMF, UPF, PCF, AUSF, NRF, UDM and NSSF on SDN principles with micro services architecture-based core network system software hosted on COTS server farms following cloud native architecture. Hardware NEs have been replaced with SDN stack, integrated NW functions have migrated to distributed NW functions, one geo location has migrated to multiple cloud-centric locations, interface-based core NW has migrated to service based architectural (SBA) change (MCP to API I/F), slice-based service delivery function, central to distributed, i.e., C-core to D-core, integrated single plane to two planes, i.e. control plane to user plane, have been the migration trends for core NW, thus fully embracing the cloud native architectural core NW layer.
Digital transformation in telco NWs involves scale of transformation in the quantum of spectrum, site throughputs, area throughputs, number of subs per cell site and core NW capacity. It is worth noting the quantum of change and scale, as given below:
- Spectrum holding per telco: 60 MHz to 1 GHz (mmWave).
- Site throughputs DL: 4 Gbps to 120 Gbps (mmWave).
- Area throughput per sq. km: 15 Gbps to 2 Tbps (mmWave).
- NW capacity in number of subs/sq. km: 26 K to 1.2 M (mmWave).
Core NW capacity is realised through the deployment of high performance computing (HPC) with a combination of multi core CPUs and GPUs with software defined storage (SDS). Core NW capacity is scalable from nx400 Gbps catering to nx10M subs in a modular capacity of enhancement. The telco core NW is thus conforming to cloud native architectures.
Telco NW data comes from multiple sources like 1. OSS /NoC; 2. CRM/CEM platform of telcos; 3. Service delivery platforms (SDPs) for social NW apps like LinkedIn/Facebook/Twitter/Google; and 4. Geo-based data sources for NW elements. All such sources have to go through the same stages of collect/clean/correlate/train and modelling, and a secondary level of modelled data goes through another phase of collect/clean/correlate/train and modelling before another algorithm-level approach is applied for decision-making in RAN NW optimisation. SON is a reality in the cloud native and AI-ML era.
Macrolevel functions of ML/AI adoption in telco cloud include the following:
- Integration of OSS data ML/AI workflow
- Predetection function
- Decision-making function with regard to threshold attainment for triggering the SON function on AI/ML on AI/ML functional module and presentation and communication module
- NPI action for optimisation
This is complemented by a hardware hosting platform for collecting data from the OSS, i.e, data handling/training farms with GPUs and inferencing platforms, i.e., refers to the process of taking a model that has already been trained and using that trained model to make useful predictions.
Migration from rule-based tool adoption to SON-based tool deployment is driven by the following attributes. 1. Fixed calculation mode to deep analytics mode using ML; 2. Limited analysis of parameters to highly scalable parameters involved in large cell dense environments; 3. Predefined decision-making based on a pre-set threshold to automated variable deduced interferences; 4. Single-point NW data source to multiple NW data sources coming from various sources such as OSS/BSS, SDP/CRM/CEM; 5. Static algorithms to self-learning algorithmic evolution model; 6. Fixed model to learned automation.
Digital transformation in network operations is for managing the telco cloud at scale and delivering superior customer experience. Cloud native intelligent NW operations are from a transactional to a self-learning automation-native environment, robotic process automation, predictive network healing and machine learning-based operational processes are for managing the massive scale of NW elements. People have to be highly skilled service engineers of a special task force of telecom NW operations, enabling the organisation for the future. Platform involves unfiltered massive data to enable intelligent operations through AI/ML. Partners are in a connected environment with ecosystem partners enabling innovation for an intelligent service delivery process and integrated business and service support operations.
Automated telco cloud operations are captured through the use of the telco cloud manager (VNFM) and the network services orchestrator (NSO) as defined in standardisation for virtualisation in the cloudification era of adoption. The objective is zero touch automation of the complex Fault, Configuration, Accountings, Performance and Security (FCAPS) procedures, which are rendered E2E in a coordinated automation mode. SDN NFV architectures as per standards-based process is enabling a massive scale of NE and task managements.
Telco cloud is a reality with proven experience and time-tested stability.
