Virtual Machine

Description

A Virtual Machine (VM) is a software-based abstraction that emulates the functionality of a physical computer, including its CPU, memory, storage, and network interfaces. It runs on a hypervisor (or Virtual Machine Monitor) installed on physical server hardware. The hypervisor allocates physical resources (compute, storage, network) to each VM, ensuring isolation and security between multiple VMs running on the same host. Each VM contains its own guest operating system and application software, behaving like an independent physical machine from the perspective of the software running inside it. This isolation is a key architectural principle, preventing faults or security breaches in one VM from affecting others on the same host.

In the context of 3GPP networks, VMs are the primary execution environment for Virtualized Network Functions (VNFs). Core network functions like the MME, SGW, PGW, and later the AMF, SMF, and UPF in 5GC, can be implemented as software packages deployed within VMs. The hypervisor manages the lifecycle of these VMs, including instantiation, scaling, migration, and termination. Resource allocation is dynamic; the hypervisor can adjust CPU cores, memory, and virtual network interface bandwidth assigned to a VM based on the VNF's current load and performance requirements, a capability managed by an NFV Orchestrator (NFVO).

The role of the VM is central to the ETSI NFV architectural framework, which 3GPP adopts and references. The VM provides the necessary isolation and resource guarantees for carrier-grade network functions. It allows operators to deploy network services on commercial off-the-shelf (COTS) hardware in data centers, moving away from proprietary, integrated appliances. This shift enables rapid service innovation, as new VNFs can be developed and deployed as software updates. Furthermore, VMs support high availability through live migration, where a running VM can be moved to another physical host with minimal service interruption for maintenance or load balancing.

Purpose & Motivation

The VM was introduced into 3GPP specifications to support the industry-wide shift towards Network Function Virtualization (NFV). Prior to NFV, telecom networks were built using monolithic, proprietary hardware appliances for each network function (e.g., a physical firewall, a physical MME). This approach led to long innovation cycles, high capital and operational expenses, vendor lock-in, and inefficient resource utilization, as each appliance was sized for peak load and often remained underutilized.

The purpose of standardizing the VM concept within 3GPP was to provide a common, interoperable foundation for virtualized deployments. It solves the problem of hardware dependency by abstracting network functions into software that can run on standardized cloud infrastructure. This enables operators to achieve greater agility, scaling services up or down elastically based on demand. It also facilitates multi-vendor environments, as VNFs from different vendors can, in principle, run on the same NFV Infrastructure (NFVI) provided they conform to the VM specifications and interfaces.

Historically, the work was driven by operator demands for cost reduction and operational flexibility, leading to collaboration between 3GPP and ETSI ISG NFV. The initial integration in Release 10 marked the beginning of formal architectural support for virtualization, paving the way for the cloud-native evolution seen in later releases with containers and microservices.

Key Features

• Hardware abstraction and isolation via a hypervisor
• Independent guest operating system per VM
• Dynamic resource allocation (vCPU, memory, storage)
• Support for live migration between physical hosts
• Carrier-grade performance and security isolation
• Standardized interfaces for management and orchestration (e.g., via VIM)

Evolution Across Releases

Defining Specifications