דילוג לתוכן הראשי

5G Networks Guide: SA vs NSA, Slicing, Spectrum and 5G-Advanced

5G Networks Guide: SA vs NSA, Slicing, Spectrum and 5G-Advanced

This 5G networks guide explains how fifth-generation mobile networks actually work — the difference between standalone and non-standalone 5G, network slicing, carrier aggregation, spectrum bands, and the 5G-Advanced evolution. 5G is the fifth generation of cellular technology, standardised by 3GPP from Release 15 (2018), delivering far higher speeds, much lower latency, and massive device density than 4G LTE, and — in its standalone form — a programmable, sliceable network rather than just a faster radio. Whether you are an operator, vendor, or enterprise buyer, this guide covers the architecture and capabilities that define modern wireless.

TL;DR — key takeaways

  • 5G NSA bolts a 5G radio onto a 4G core (fast to deploy); 5G SA uses a cloud-native 5G core end to end and unlocks slicing, ultra-low latency, and advanced QoS.
  • Network slicing runs many isolated virtual networks on one physical 5G infrastructure, each tuned to a use case.
  • Carrier aggregation combines multiple frequency blocks for higher throughput.
  • Spectrum: low-band for coverage, mid-band (C-band) for capacity, mmWave for dense hotspots.
  • 5G-Advanced (Release 18, 2024) adds AI, energy efficiency, XR, and improved IoT (RedCap/eRedCap).

What is 5G, and how is it different from 4G?

5G is the fifth generation of mobile network technology, standardised by 3GPP beginning with Release 15 in 2018. Compared with 4G LTE it delivers dramatically higher peak and average speeds, much lower latency (single-digit milliseconds in ideal standalone deployments), and the ability to support a far greater density of connected devices per square kilometre. But the deeper difference is architectural: 5G is designed to be a programmable, software-defined, cloud-native network. In its full standalone form, 5G is not just a faster pipe but a platform that can be sliced into many purpose-built virtual networks, integrated with edge computing, and offered to enterprises with guaranteed service levels. That is why the industry frames 5G as an enabler of new use cases — industrial automation, fixed wireless broadband, immersive media, and mission-critical connectivity — rather than merely a generational speed bump.

5G SA vs NSA: why standalone matters

5G comes in two deployment flavours. Non-Standalone (NSA), defined in Release 15, attaches a 5G New Radio to an existing 4G LTE core (the EPC): the 4G network handles control signalling while the 5G radio adds data capacity and speed. NSA was the fast on-ramp because it reused 4G investment, but it cannot deliver 5G’s full feature set. Standalone (SA) uses a brand-new, cloud-native 5G core (5GC) end to end with no 4G dependency. Only SA unlocks the transformative capabilities — ultra-low latency, network slicing, massive machine-type communications, and advanced quality-of-service — because those features are implemented in the 5G core, not the radio. The global migration from NSA to SA is the central story of the 5G era; until an operator runs SA, it is essentially offering faster 4G rather than true 5G.

Aspect5G NSA5G SA
Core networkReuses 4G EPCNew cloud-native 5G core (5GC)
Control planeVia 4G LTENative 5G
Network slicingNoYes
Ultra-low latencyLimitedYes
Deployment speedFaster (reuses 4G)Requires new core
3GPP originRelease 15 (2018)Release 15+ (full SA)

What is network slicing and why is it commercially important?

Network slicing lets an operator run multiple logical, end-to-end virtual networks over one shared physical 5G infrastructure, each slice tuned to a distinct requirement and isolated from the others. A low-latency, high-reliability slice can serve industrial automation or remote control; a massive-connectivity slice can carry millions of IoT sensors; a high-throughput slice can handle live video. Slicing spans the radio, transport, and core, and depends on the 5G standalone core together with virtualisation and orchestration. Its commercial importance is that it converts one network into many purpose-built ones without separate hardware, letting operators sell differentiated, SLA-backed connectivity to enterprises — effectively offering a private-grade network experience as a service. Slicing, more than raw speed, is the basis of many of the new enterprise revenue models that justify 5G standalone investment.

Carrier aggregation and spectrum bands

Carrier aggregation (CA) combines multiple frequency blocks into one wider effective channel, so a device gets more bandwidth and higher throughput than any single block could provide — for example aggregating a high-capacity high-band carrier with a far-reaching low-band carrier to get both speed and coverage at once. CA exists across bands and even across 4G and 5G carriers in NSA. The spectrum it aggregates falls into three ranges with different trade-offs: low-band (below ~1 GHz) travels far and penetrates buildings, giving nationwide coverage at modest speed; mid-band (roughly 1–7 GHz, including the pivotal C-band near 3.5 GHz) balances coverage and capacity and carries most of the real 5G speed gain; and high-band millimetre wave (mmWave, ~24 GHz+) offers huge capacity over very short range for dense hotspots and fixed wireless access. Operators blend all three, using low-band for reach, mid-band for bulk capacity, and mmWave for targeted density.

What is 5G-Advanced?

5G-Advanced is 5G from 3GPP Release 18 onward, the second major phase of 5G evolution after Releases 15–17. Release 18 was frozen in 2024. It enhances existing capabilities and adds new ones: deeper integration of AI and machine learning into the air interface and network management, significant network energy-efficiency improvements, enhanced positioning, better support for extended reality (XR) and immersive media, expanded RedCap/eRedCap for IoT, uplink and MIMO enhancements, and refinements to non-terrestrial (satellite) networks. 5G-Advanced is the bridge between original 5G and eventual 6G, delivering substantial capability gains on the installed 5G standalone base. For operators it is an upgrade path that extracts more value and unlocks new use cases from existing 5G investments rather than demanding a wholesale new generation — making it a central planning theme for the second half of the 2020s.

RedCap: extending 5G to mid-tier devices

RedCap (Reduced Capability) is a 5G NR device class introduced in Release 17 (2022) for devices that need more than low-power IoT (NB-IoT/LTE-M) but less than a smartphone. By cutting bandwidth, antenna count, and complexity, RedCap lowers device cost, power draw, and size while still delivering far higher throughput and lower latency than 4G LPWA — suiting industrial sensors, wearables, and video surveillance. Its key contribution is letting these mid-tier devices connect natively to a 5G standalone network instead of relying on 4G, so an operator can serve the whole device spectrum — smartphones, RedCap mid-tier, and LPWA sensors — on a single 5G network. Release 18 adds the even leaner eRedCap variant. RedCap is how 5G broadens beyond phones into the large middle market of connected devices, and it pairs naturally with network slicing and edge computing for industrial deployments.

Related on DROAM

Frequently asked questions

What is the difference between 5G SA and 5G NSA?

5G NSA (Non-Standalone) and 5G SA (Standalone) are two ways to deploy 5G. NSA, the first wave defined in 3GPP Release 15 (2018), bolts a 5G radio (NR) onto an existing 4G LTE core (the EPC), using the 4G network for control signalling and the 5G radio mainly for extra data capacity and speed. It was fast to deploy because it reused the 4G core, but it cannot deliver 5G’s full feature set. 5G SA uses a new, cloud-native 5G core (5GC) end to end, with no 4G dependency. Only SA unlocks the headline 5G capabilities — ultra-low latency, network slicing, massive device density, and advanced QoS — because those features live in the 5G core. The industry is steadily migrating from NSA to SA, which is the architecture that makes 5G genuinely transformative rather than just faster 4G.

What is 5G network slicing?

Network slicing lets an operator run multiple logical, end-to-end virtual networks on top of one shared physical 5G infrastructure, each “slice” tuned to a different requirement and isolated from the others. One slice might guarantee ultra-low latency and high reliability for industrial automation or remote control; another might provide massive connection density for IoT sensors; another high throughput for video. Slicing spans the radio, transport, and core, and depends on the 5G standalone core plus virtualisation and orchestration. It matters because it turns a single network into many purpose-built ones without separate hardware, enabling operators to offer differentiated, SLA-backed connectivity to enterprises — selling, in effect, a private-grade network experience as a service. Slicing is one of the defining commercial opportunities of the 5G SA era.

What is carrier aggregation in 5G?

Carrier aggregation (CA) combines multiple frequency blocks (component carriers) into a single, wider effective channel so a device gets more bandwidth and higher throughput than any one block could provide. A phone might aggregate a wide-but-short-range high-band carrier with a narrower-but-far-reaching low-band carrier, getting both the speed of the high band and the coverage of the low band simultaneously. Carrier aggregation existed in 4G LTE and is extended in 5G NR, including aggregating across bands and even across 4G and 5G carriers in NSA deployments. It is one of the principal tools operators use to deliver gigabit speeds from spectrum that is otherwise fragmented across many separate bands, and it is fundamental to getting the most out of an operator’s licensed spectrum holdings.

What is 5G-Advanced?

5G-Advanced is the name for 5G from 3GPP Release 18 onward, marking the second major phase of 5G evolution (the first being Releases 15–17). Release 18 was frozen in 2024. It enhances the existing 5G capabilities and adds new ones: deeper AI/machine-learning integration into the air interface and network management, improvements to network energy efficiency, enhanced positioning, better support for extended reality (XR) and immersive media, expanded RedCap/eRedCap for IoT, improved uplink and multiple-input multiple-output (MIMO), and refinements to non-terrestrial (satellite) networks. 5G-Advanced is the bridge between the original 5G and the eventual 6G, delivering meaningful capability gains on the installed 5G base. For operators it is an upgrade path that extracts more value and new use cases from existing 5G standalone investments rather than requiring a new generation.

What are the main 5G spectrum bands?

5G uses three broad spectrum ranges, each with a different coverage-versus-capacity trade-off. Low-band (below ~1 GHz) travels far and penetrates buildings well, giving wide coverage but modest speeds — it is the foundation for nationwide 5G. Mid-band (roughly 1–7 GHz, including the pivotal C-band around 3.5 GHz) is the workhorse of 5G, balancing good coverage with substantial capacity, and is where most of the meaningful 5G speed improvement lives. High-band, or millimetre wave (mmWave, ~24 GHz and above), offers enormous capacity and the fastest speeds but very short range and poor obstacle penetration, so it is used for dense hotspots like stadiums, venues, and fixed wireless access. Operators combine all three, typically using low-band for reach, mid-band for the bulk of capacity, and mmWave for targeted high-density coverage.

What is RedCap and how does it relate to 5G?

RedCap (Reduced Capability) is a 5G NR device class introduced in 3GPP Release 17 (2022) for devices that need more than low-power IoT (NB-IoT/LTE-M) but less than full 5G. By reducing bandwidth, antenna count, and complexity, RedCap lowers device cost, power, and size while still offering meaningfully higher throughput and lower latency than 4G LPWA — targeting industrial sensors, wearables, and video surveillance. Crucially, RedCap lets these mid-tier devices connect natively to a 5G standalone network rather than relying on 4G, so operators can serve the full IoT spectrum on one 5G network. Release 18 adds eRedCap, an even lower-capability variant. RedCap is how 5G extends its reach into the broad middle of the device market, complementing smartphones at the top and IoT sensors at the bottom.