4G/LTE vs 5G: What They Are and How They Differ

4G LTE and 5G are both standardised by 3GPP — the same technical body — but they represent distinct generations with different core architectures, spectrum strategies, and design goals. Understanding the key differences matters when choosing a SIM plan, evaluating coverage, or deciding whether a 5G device provides a practical benefit in a given location.

The key points covered in this article:

• What 4G LTE is and how it differs technically from 3G predecessors
• How 5G NR is structured, including the critical difference between NSA and SA deployments
• How Sub-6 GHz and mmWave 5G serve different use cases
• What the 3GPP Release system means and which Releases introduced each generation
• How 4G and 5G handle speed, latency, and connection density differently
• How voice calls work across both generations (VoLTE and VoNR)

For more on how voice calls work on LTE and 5G networks, see VoLTE and VoNR Explained. For definitions of terms used in this article, see the SIM & Mobile Glossary.

The table below gives an orientation before the detailed sections.

What Is 4G LTE?

4G refers to the fourth generation of mobile network standards. The ITU-R defined the requirements for 4G under the IMT-Advanced framework, which operators and standards bodies worked to satisfy through the 2000s and 2010s.

LTE stands for Long-Term Evolution. It is the specific radio access technology that became the dominant implementation of 4G. LTE was standardised in 3GPP Release 8, which was frozen in December 2008. Release 8 defined LTE as a purely packet-switched network — meaning all traffic, including data and eventually voice, is carried as IP packets. There is no circuit-switched voice plane in LTE, which is why VoLTE (the IMS-based system for carrying voice over LTE) had to be developed separately.

The 3G predecessor, UMTS/HSPA (defined in 3GPP Releases 5–7), carried voice over a circuit-switched domain and data over a packet-switched domain. LTE eliminated the circuit-switched domain entirely, representing a fundamental architectural change rather than an incremental evolution.

LTE-Advanced and LTE-Advanced Pro

The base LTE specification was extended in two significant later Releases:

• LTE-Advanced (Release 10, 2011): Introduced carrier aggregation (CA), which allows a device to combine multiple frequency bands simultaneously to increase throughput. Release 10 also introduced enhanced MIMO (Multiple Input, Multiple Output) antenna configurations and added relay nodes to improve coverage.
• LTE-Advanced Pro (Release 13, 2016): Added features including licensed-assisted access (LTE-U/LAA), narrowband IoT (NB-IoT), enhanced carrier aggregation, and support for up to 32 component carriers. Marketing terms such as “4.5G” and “Gigabit LTE” typically refer to LTE-Advanced Pro devices and networks.

These extensions kept LTE technically competitive through the late 2010s while 5G was under development.

What Is 5G NR?

5G NR (New Radio) is the radio access technology of the fifth generation, standardised in 3GPP Release 15. The NSA (Non-Standalone) variant of Release 15 was frozen in December 2017; the SA (Standalone) variant’s physical layer was frozen in June 2018, with the ASN.1 (protocol encoding) freeze completed in September 2018.

5G NR covers two frequency ranges defined in 3GPP:

• FR1 (Sub-6 GHz): Frequencies from 410 MHz to 7125 MHz. This range includes re-farmed low-band spectrum (600–900 MHz), mid-band spectrum widely used for 5G capacity (2.5–3.7 GHz, including the internationally co-ordinated n77/n78/n79 bands), and extended mid-band up to 7.125 GHz.
• FR2 (mmWave): Frequencies from 24.25 GHz to 52.6 GHz. This range enables very high bandwidth but with limited propagation distance and sensitivity to physical obstructions.

5G NR uses Orthogonal Frequency-Division Multiplexing (OFDM) as its underlying modulation scheme, the same approach as LTE. However, 5G NR introduces a more flexible subcarrier spacing (numerology) that allows the same physical layer to operate efficiently across both low-latency mmWave links and wide-coverage Sub-6 GHz links.

NSA vs SA: The Architecture Divide

The distinction between NSA (Non-Standalone) and SA (Standalone) is among the most practically important aspects of 5G deployments.

Non-Standalone (NSA)

In NSA mode, 5G NR is deployed as an additional radio access layer on top of an existing 4G LTE Evolved Packet Core (EPC). The LTE cell provides the control plane (signalling, session management, mobility), while the 5G NR cell adds a second data plane for higher throughput. This architecture is defined in 3GPP as Option 3 (EN-DC: E-UTRA/NR Dual Connectivity).

Practical consequences of NSA:

• The device must maintain an LTE connection at all times. If LTE signal is lost, the 5G NR connection also falls.
• Voice calls continue to use VoLTE on the LTE layer. 5G NR does not handle voice in NSA mode.
• Network slicing and URLLC (ultra-reliable low-latency communications) features require 5G SA and are not available in NSA.
• NSA was the primary deployment mode for the first wave of commercial 5G networks from 2019 onward because it allowed carriers to add 5G NR capacity without replacing the LTE core.

Standalone (SA)

In SA mode, 5G NR connects directly to a 5G Core (5GC), which is an entirely new core network architecture defined in 3GPP Release 15. The 5GC uses a Service-Based Architecture (SBA) with new network functions: the AMF (Access and Mobility Management Function) replaces the LTE MME, the SMF (Session Management Function) replaces PGW/SGW, and the UPF (User Plane Function) handles forwarding.

SA enables:

• VoNR (Voice over New Radio): voice calls carried natively over 5G NR via IMS
• Network slicing: logical partitioning of the network for different traffic types or customers
• URLLC (ultra-reliable low-latency communications): designed for industrial and mission-critical applications
• Full independence from LTE infrastructure, allowing carriers to eventually retire LTE spectrum

As of 2026, 5G SA networks are operational in a growing number of markets, including South Korea, China, parts of the United States, and Japan, but LTE-anchored NSA coverage continues to represent the majority of 5G connections globally. The transition to SA is carrier-specific: operators that have invested in 5G SA cores can offer VoNR and network slicing, while operators still on NSA 5G cannot, regardless of the device’s hardware capability.

From a consumer perspective, a practical way to identify whether you are on 5G SA is to check whether VoNR is active. On Android, field test modes that show IMS registration may indicate VoNR registration separately from VoLTE. On iPhone, iOS does not expose VoNR status directly; VoNR operation on supported carriers is handled automatically.

Sub-6 GHz vs mmWave: Two Very Different 5G Experiences

The two frequency ranges used for 5G NR have fundamentally different physical characteristics, leading to distinct deployment strategies.

Sub-6 GHz 5G

Sub-6 GHz 5G behaves similarly to LTE in terms of coverage: the signal propagates over kilometres from a single site, penetrates buildings, and provides broad area coverage. Low-band 5G (600–900 MHz) covers the widest areas but has lower peak throughput. Mid-band 5G (2.5–4.2 GHz, especially the 3.5 GHz band standardised as n77/n78) provides the best balance of coverage and throughput and is the workhorse of most national 5G deployments.

For most consumers, the 5G connection they experience daily is Sub-6 GHz mid-band. The practical throughput improvement over LTE-Advanced depends on network loading, spectrum allocation, and device capability, but mid-band 5G consistently delivers higher median throughput than LTE in controlled conditions.

mmWave 5G (FR2)

mmWave 5G uses frequencies from roughly 24 GHz to 52.6 GHz. Electromagnetic waves at these frequencies have very short wavelengths that are absorbed by building materials, trees, and even rainfall. A typical mmWave small cell covers an area of hundreds of metres rather than kilometres. Line-of-sight or near-line-of-sight conditions are generally required for reliable connection.

The peak throughput achievable on mmWave significantly exceeds Sub-6 GHz because the available bandwidth at mmWave frequencies is much larger. However, consistent mmWave coverage in a large geographic area is not economically or physically feasible, so mmWave deployment is concentrated in specific high-density environments: major transit hubs, stadiums, convention centres, and fixed wireless access for homes and businesses.

Consumers in most markets will rarely encounter mmWave 5G in everyday use. A device’s 5G specification sheet distinguishes FR1-only and FR1+FR2 support; the latter adds mmWave antenna modules that increase device cost and complexity.

Speed, Latency, and Connection Density

The three primary design objectives that 3GPP defined for 5G NR under the IMT-2020 framework differ from those of 4G LTE in emphasis and scale. IMT-2020 defined three use case categories for 5G: eMBB (Enhanced Mobile Broadband), URLLC (Ultra-Reliable Low-Latency Communications), and mMTC (Massive Machine-Type Communications). Most consumer 5G services focus on eMBB; URLLC and mMTC are primarily relevant for industrial and IoT deployments.

Throughput

LTE (Release 8) was specified with a theoretical peak downlink of 300 Mbps under ideal conditions (4×4 MIMO, 20 MHz bandwidth). LTE-Advanced extended this through carrier aggregation. 5G NR can aggregate much wider bandwidths — especially at mmWave — leading to higher theoretical peaks. However, real-world throughput on both technologies depends on spectrum allocation, cell loading, device capability, and distance from the base station.

This article does not cite specific speed figures because real-world performance varies significantly by market, operator, and location. Independent measurement organisations such as Opensignal and Ookla publish country-level benchmarks; searching for the specific operator’s name alongside “5G speed test” will return recent measurement data for a given market.

Latency

LTE has a theoretical user-plane latency specification in the single-digit milliseconds range under ideal conditions. 5G NR was designed with additional latency reduction targets, particularly for URLLC applications in 5G SA. For the vast majority of consumer applications — web browsing, video streaming, messaging — the latency difference between LTE and mid-band 5G SA is not perceptible in practice. URLLC latency targets are relevant for industrial automation, remote surgery, and vehicle-to-everything (V2X) communications, which require consistent low latency guarantees that cannot be met with best-effort scheduling.

Connection Density

LTE was designed with a connection density target that became a bottleneck in dense event venues and transit hubs under high load. 5G NR’s IMT-2020 specification increases the target connection density per unit area, and the flexible numerology of 5G NR allows the system to handle many short transmissions efficiently. The mMTC use case specifically targets large numbers of low-power IoT devices per cell. For consumers, this translates to more stable connections in stadiums, conference centres, and busy transport hubs where LTE networks can become congested.

MIMO and Beamforming

LTE introduced MIMO antenna techniques, with practical deployments reaching 4×4 MIMO on mid-band LTE-Advanced networks. 5G NR introduces Massive MIMO, where base station antenna arrays can contain dozens to hundreds of antenna elements. Combined with beamforming — shaping the radio signal toward specific devices rather than broadcasting in all directions — Massive MIMO improves spectral efficiency and allows multiple users to be served on the same frequency resource simultaneously. This is particularly effective in mid-band 5G deployments using bands such as n77/n78, where antenna arrays of 64 or more elements are commonly deployed.

3GPP Releases: The Versioning System Behind Each Generation

3GPP defines mobile network specifications in numbered Releases, each representing a functionally stable snapshot of the specification set. New features are always tied to a specific Release.

Understanding Release numbers is relevant when evaluating device specifications. A device described as supporting “5G” may implement Release 15 features only, while a later device may implement Release 16 or Release 17 features that add measurable improvements for specific use cases. When a carrier advertises a network feature — such as network slicing or satellite connectivity — checking which 3GPP Release introduced that feature will clarify whether a given device supports it. Device specification sheets often list supported 3GPP Release or explicitly list supported features by name.

How Voice Calls Work Across Both Generations

Voice call handling is one of the clearest practical differences between 4G, 5G NSA, and 5G SA deployments.

On 4G LTE, all voice calls use VoLTE (Voice over LTE) — carried as IP packets over the LTE network via an IMS (IP Multimedia Subsystem) core. The 3G circuit-switched fallback (CSFB) that earlier LTE devices used has been eliminated in all markets that have completed their 3G shutdown.

On 5G NSA, the device maintains an LTE anchor for the control plane. Voice calls still use VoLTE on the LTE layer. The 5G NR radio does not carry voice in NSA mode.

On 5G SA, voice calls use VoNR (Voice over New Radio), standardised in GSMA NG.114 and 3GPP Release 15. VoNR routes the call directly over the 5G NR bearer via IMS, using the same IMS infrastructure as VoLTE but with a 5G-native bearer and support for the EVS (Enhanced Voice Services) codec, which delivers higher audio quality than the AMR-WB codec used in VoLTE. For devices on a 5G SA network without VoNR support, fallback to VoLTE on LTE is the standard behaviour.

For the complete technical explanation of IMS, SIP signalling, QoS bearers, and what happens when VoLTE is not supported, see VoLTE and VoNR Explained.

What This Means When Choosing a SIM

The distinction between 4G and 5G, and between 5G NSA and SA, has practical consequences when selecting a carrier or plan.

Network coverage layer. A carrier’s 5G coverage map reflects where 5G NR is deployed, but 4G LTE remains the coverage layer in most areas. Evaluating a carrier’s LTE coverage is at least as important as 5G availability, particularly for travellers and users in suburban or rural locations. For more on how MNOs and MVNOs differ in network access, see What Is an MNO?.

SIM and eSIM compatibility. 5G capability depends on the device, not the SIM card. A 5G-capable device on a 4G-only SIM plan will connect at 4G speeds. Conversely, a 4G-only device with a 5G plan will connect at 4G speeds. The SIM provides authentication and plan entitlements; the network selects the highest radio technology the device supports. For more on SIM vs eSIM, see SIM vs eSIM.

Data efficiency. 5G does not reduce the data volume your applications consume. If you are on a limited data plan, a faster connection will deplete your allowance more quickly unless you use data-saving settings. For practical tips on reducing mobile data use, see Data-Saving Techniques.

Roaming. When roaming internationally, your device connects to the visited network’s technology. A 5G device will use 4G LTE on a visited network that has not yet enabled 5G roaming agreements. Roaming data speeds depend on the visited network’s LTE or 5G capability and the roaming agreement between your home carrier and the visited operator. See International Roaming for how roaming agreements and data usage work across borders.

MVNOs and 5G. MVNOs lease network capacity from MNOs and typically pass through the host network’s radio technology, including 5G, where the wholesale agreement permits it. Whether an MVNO’s plan includes 5G access depends on the specific agreement with the host MNO. Some MVNOs offer 5G-capable plans; others remain limited to 4G LTE even on a host network that has 5G. Confirm 5G availability with the specific MVNO before selecting a plan if 5G is a priority. See What Is an MNO? for how the MNO–MVNO wholesale relationship works.

Device band compatibility. Not all 5G devices support all 5G bands. A device sold in one market may lack the n78 (3.5 GHz) band widely used for 5G in another market, or the specific low-band 5G frequency a carrier uses. When using an unlocked device internationally, checking band compatibility is as important for 5G as it is for 4G LTE.

You can filter and compare SIM plans by network generation on SimFinder.

FAQ

See the frontmatter above for structured FAQ entries compatible with schema.org/FAQPage. The questions below cover additional topics.

When did commercial 5G first launch? The first commercial 5G NR service launched in South Korea on 3 April 2019, when SK Telecom, KT Corporation, and LG Uplus activated 5G NSA service simultaneously. The United States saw its first commercial 5G deployments from Verizon (fixed wireless, October 2018) and AT&T and T-Mobile (mobile 5G NR, late 2018–2019). These dates refer to the first commercial availability; widespread consumer coverage developed over subsequent years.

Can a 4G SIM card be used on a 5G device? Yes. A 4G SIM provides authentication credentials that are valid on any network generation the carrier and device support. The SIM authenticates the subscriber; the radio access technology is negotiated between the device modem and the base station. However, some carriers issue updated SIMs or eSIM profiles to support specific 5G SA features. If a carrier requires an updated profile for 5G SA features, the carrier’s support page will confirm this.

What does “5G UC”, “5G+”, or “5G UW” mean on a phone’s status bar? These are carrier-specific indicators for different 5G capability tiers. “5G UW” (Ultra Wideband, Verizon) and “5G+” (AT&T) indicate mmWave or high-band mid-band connections. “5G UC” (Ultra Capacity, T-Mobile) indicates mid-band 5G (primarily n41 2.5 GHz). Plain “5G” typically indicates low-band 5G. These labels are branding decisions by individual carriers, not standardised 3GPP terminology.

Is 5G available when roaming internationally? 5G roaming availability depends on whether both the home carrier and the visited network have established 5G roaming agreements and whether the visited network has deployed compatible 5G bands. As of 2026, 5G roaming agreements are expanding but are not universal; many roaming connections default to 4G LTE. The specific countries and bands supported for roaming are listed on each carrier’s roaming information page. See International Roaming for how roaming agreements work.

What is 5G Standalone (SA) network slicing? Network slicing is a 5G SA feature defined in 3GPP Release 15 that allows a carrier to create multiple logically isolated virtual networks on the same physical infrastructure. Each slice can have different QoS parameters — for example, one slice with guaranteed low latency for emergency services and another optimised for high-throughput consumer data. Network slicing requires the 5G Core’s NSSAI (Network Slice Selection Assistance Information) framework and is not available on 5G NSA or LTE networks.

Related Guides

• VoLTE and VoNR Explained — How voice calls travel over 4G and 5G networks via IMS, and what VoNR requires from a 5G SA core
• What Is an MNO? — How mobile network operators hold spectrum licences and build the infrastructure that 4G and 5G run on
• SIM vs eSIM — How the SIM authenticates you to the network regardless of whether you are on 4G or 5G
• International Roaming — How 4G and 5G roaming agreements determine which network technology is available abroad
• Data-Saving Techniques — How to manage data consumption on faster 5G connections