4G/5G cellular networks

4G and 5G are generations of cellular networks ("G" stands for generation.)

4G download speeds are around 20 Mbps or higher.

Specifications:

• Internet protocols are defined in RFCs
• 4G and 5G networks are defined in Technical Specifications

Cellular communications

The electromagnetic spectrum has only a limited number of channels, and many users compete for them.

Cellular communication solves this limitation by dividing coverage areas into smaller regions called cells:

• each cell has a base station (cell tower) that sends and receives signals from mobile devices within that area
• carriers place many antennas across the landscape so they can reuse the same frequencies in different nonadjacent cells
• cells are often represented as a idealized hexagonal grid, but their actual shape and size vary depending on real-world conditions

The coverage area of a cell depends on many factors:

• the transmitting power of the base station and mobile devices
• obstacles like buildings or mountains
• the height and type of the base station antennas
• the number of users and demand for service in the area

A cellular carrier (roughly opérateur in French) is a company that provides mobile phone service in a geographic area.

4G LTE: architecture

4G LTE networks implement the LTE (Long-Term Evolution) radio access technology standardized by 3GPP.

They are architecturally divided into two parts:

1. the radio access network (E-UTRAN) at the cellular edge, which provides the radio link between mobile devices and base stations (the first hop into the network)
2. the Evolved Packet Core (EPC), an IP-based core network that manages connectivity, mobility, authentication, and data routing between:
   • the carrier's own network
   • other carrier networks
   • the Internet

4G LTE: elements

Major elements of the 4G LTE network architecture:

Component	Description
Mobile Device (UE - User Equipment)	A device connecting to a cellular network (e.g., smartphone, tablet, laptop, IoT device) containing: IP address: assigned by the P-GW (from a carrier-managed address pool) SIM card: stores subscriber service information and encryption keys IMSI (International Mobile Subscriber Identity): uniquely identifies the subscriber and the home network
Cellular Base Station (eNode-B)	The network edge element that manages devices and radio resources within its cell: allocates radio resources, schedules uplink/downlink transmissions, manages handovers communicates with other base stations to: support mobility between cells coordinate radio resource usage to reduce interference between neighboring cells In LTE terminology, it is called eNode-B: e = evolved, referring to the evolution of the 3G Node B Node B was the term used for 3G radio base stations in 5G systems the term is "ng-eNB" (this name opaqueness shows no signs in stopping)
Home Subscriber Server (HSS)	A control-plane database containing subscriber information for the carrier's home network. Used with the MME for device authentication.
Gateway Routers (S-GW and P-GW)	Serving Gateway (S-GW): routes data between base stations and the P-GW often collocated with the P-GW Packet Data Network Gateway (P-GW): connects LTE to the Internet assigns IP addresses and may perform NAT before traffic exits to the Internet last LTE element before traffic reaches the Internet
Mobility Management Entity (MME)	A control-plane element responsible for authentication, mobility, and tunnel management: acts as a middleman between the mobile device and the HSS: when a device is connected to its home network: the MME contacts the HSS within the same network when a device is roaming: the visited network's MME contacts the HSS in the device's home network controls the setup of LTE data tunnels used for forwarding traffic and supporting mobility tracks the location of active devices: base stations report device movement between cells to the MME if a device is sleeping (idle mode), the MME tracks the device by its tracking area and initiates paging (wakeup) through the relevant base stations

4G LTE: tunnels

The data path from the mobile device to the carrier gateway consists of:

1. a wireless first hop between the mobile device and base station (UE ↔ eNode-B)
2. concatenated IP tunnels:
   • between the base station and S-GW (eNode-B ↔ S-GW)
   • between the S-GW and P-GW (S-GW ↔ P-GW)

Tunnels are established under MME control when the mobile device first attaches to the network:

• each tunnel endpoint is identified by a Tunnel Endpoint Identifier (TEID)
• when the base station receives datagrams from the mobile device:
  • it encapsulates them using the GPRS Tunneling Protocol including the TEID
  • and sends them in UDP segments to the Serving Gateway (at the other end of the tunnel)

When a device moves:

• only the tunnel between the base station and S-GW changes
• other tunnel endpoints and their Quality of Service (QoS) settings remain unchanged

LTE protocols stacks

Since 4G LTE is an all-IP architecture:

• many of its protocols are already familiar: IP, TCP, UDP, and application layer protocols
• but there are also specific protocols:
  • at the link layer
  • at the physical layer
  • in mobility management

LTE data-plane protocol stacks at the mobile node, the base station and the serving gateway:

Most LTE-specific data-plane protocol activity occurs on the wireless link between the mobile device and the base station

The link layer on the mobile device is divided into three sublayers:

1. Packet Data Convergence Protocol (PDCP)
   • sits just below IP
   • compresses IP headers to reduce transmitted data
   • encrypts and decrypts IP datagrams
2. Radio Link Control (RLC) Protocol
   • splits large IP datagrams into smaller pieces for transmission and reassembles them at the receiver
   • provides link-layer reliable delivery using an ACK/NAK-based ARQ protocol
3. Medium Access Control (MAC)
   • coordinates access to radio resources through scheduling

LTE radio resource allocation

LTE uses Orthogonal Frequency Division Multiplexing (OFDM) on the downlink, which combines:

• frequency division multiplexing (FDM)
• time division multiplexing

Orthogonal means the signals on closely spaced frequency channels are designed to interfere very little with each other.

Each active mobile device is allocated:

• one or more 0.5 ms time slots
• on one or more frequency channels

Allocating more time slots (whether on the same or on different frequencies) increases the device's data rate.

The network can:

• reassign time slots as often as every 1 ms
• adjust the modulation scheme to change transmission speed

LTE scheduling

The LTE standard does not specify how time slots are assigned.

Instead, scheduling decisions (such as which device transmits on which frequency and at what time) are made by algorithms implemented by the LTE equipment vendor and/or network operator.

A common scheduling strategy is opportunistic scheduling, in which the base station makes efficient use of the wireless medium by:

• adapting transmissions to current wireless channel conditions
• selecting users with the best channel quality for transmission

Scheduling decisions can also consider:

• user priority
• contracted levels of service (e.g., silver, gold, or platinum)

LTE network attachment

When a mobile device connects to an LTE network, the process happens in three phases:

Step	Description
1. Attachment to a base station	the mobile device starts a bootstrap process to find and connect to a nearby base station it searches all channels and frequency bands for a primary synchronization signal broadcast by the base station every 5 ms after finding this signal, the device: stays on that frequency finds a secondary synchronization signal using information from the secondary signal, the device learns: channel bandwidth channel configurations carrier information of that base station the device then: selects a base station (preferably its home network) establishes a control-plane signaling connection across the wireless hop with that base station this connection is used during the rest of the attachment process
2. Mutual authentication	the base station contacts the MME (Mobility Management Entity) to verify the identity of both the mobile device and the LTE network after authentication: the MME knows the device and the connected base station the MME can begin setting up the data path to the Internet
3. Data path configuration	the MME contacts: the PDN gateway, which provides the device with an IP address the Serving gateway the base station these components create tunnels that allow the mobile device to send and receive IP packets through the LTE network and access the Internet

LTE power management

LTE devices use sleep modes to reduce battery consumption while maintaining connectivity.

A sleeping LTE device can be in two states:

1. discontinuous reception state:
   • entered after a few hundred milliseconds of inactivity
   • the mobile device and base station schedule regular wake-up times
   • during these times, the device monitors the channel for downstream (base station to mobile device) transmissions
   • between wake-up periods, the device radio is turned off to save power
2. idle state:
   • entered after a longer period of inactivity
   • the device releases its active radio connection and wakes up less frequently to monitor the network
   • when it wakes up, it may need to reconnect with a base station, especially if it has moved to a new area

The global cellular network: a network of networks

The global cellular network is a network of networks, similar to the Internet:

• a smartphone connects to a base station, which connects it to the user's home network
• the home network is operated by a cellular carrier
• the home network connects to:
  • other cellular carrier networks
  • the global Internet
• mobile networks interconnect with each other using:
  • gateway routers
  • private carrier networks such as IPX (especially for services like international roaming)

5G cellular networks

5G usually refers to 5G NR (New Radio), the radio access technology standardized by 3GPP.

The 5G NR air interface is not backward-compatible with 4G LTE, meaning carriers must invest heavily in new infrastructure during the transition.

5G network design focuses on three major service categories:

1. eMBB (Enhanced Mobile Broadband):
   • provides higher bandwidth and faster data rates
   • supports peak rates up to 10 Gbps under ideal conditions
   • enables applications such as HD video streaming
2. URLLC (Ultra Reliable Low-Latency Communications):
   • targets extremely low latency (around 1 millisecond)
   • ideal for mission-critical latency-sensitive applications such as factory automation and autonomous vehicles
3. mMTC (Massive Machine Type Communications):
   • targets large numbers of low-power connected devices
   • enables applications such as sensing, metering, and monitoring in IoT environments

5G frequencies

5G frequencies are divided into two groups:

1. FR1 (410 MHz–7.125 GHz)
   • most early deployments use FR1 frequencies
2. FR2 (24.25 GHz–52.6 GHz)
   • also known as millimeter wave (mmWave) frequencies
   • enables much higher data speeds because it supports wider channel bandwidths

However, FR2 has two major limitations:

1. shorter range:
   • requires denser deployments of base stations in urban areas
   • generally unsuitable for rural areas
2. greater sensitivity to interference:
   • rain, foliage, buildings, and other environmental factors can significantly reduce signal strength

5G millimeter wave frequencies

Many of the highest-capacity 5G deployments rely on millimeter wave frequencies in the 24–52 GHz range.

These frequencies can potentially increase network capacity by up to 100 times compared with 4G.

Network capacity can be described as:

capacity = cell density × available spectrum × spectral efficiency

Where:

• cell density = number of cells per km²
• available spectrum = amount of frequency bandwidth available (Hz)
• spectral efficiency = how efficiently a cell uses its available frequency to communicate with users (bps/Hz/cell)

Multiplying these units gives a capacity in bps/km2.

In 5G, all three factors increase compared with 4G:

Factor	Explanation
Higher cell density	millimeter wave signals experience higher propagation losses and poorer penetration than 4G LTE frequencies as a result, their coverage area is smaller and more base stations are needed this increases the number of cells per km²
More available spectrum	5G FR2 uses a much wider frequency range than 4G LTE (up to about 2 GHz of available spectrum) more spectrum allows more data to be transmitted at the same time
Higher spectral efficiency	information theory (Shannon's capacity limit) shows that doubling spectral efficiency requires a 17-fold increase in power instead of simply increasing power, 5G improves efficiency using MIMO (Multiple Input Multiple Output) technology: base stations use multiple antennas beamforming directs signals toward specific users instead of broadcasting in all directions spatial multiplexing allows multiple users to communicate simultaneously using the same frequency band

By increasing all three terms in the capacity equation, 5G can achieve up to a 100× increase in capacity in dense urban areas.

The wider frequency bands also allow peak download speeds of 1 Gbps or more.

However, millimeter wave signals have limited range because they are easily blocked by obstacles such as buildings and trees. To maintain reliable coverage, 5G networks need many small cells to fill gaps between users and larger base stations. In densely populated areas, these small cells may need to be deployed only tens to hundreds of meters apart.

5G core network

The 5G Core network is the central part of the 5G system responsible for managing connectivity, mobility, authentication, policy control, and data services. It supports services such as mobile Internet access and voice services.

Compared with 4G, the 5G Core:

• is built around Control Plane and User Plane Separation (CUPS)
• uses software-based, cloud-native network functions

The control plane vs data plane distinction at the network layer and CUPS in 5G Core are based on the same principle:

• control plane = makes decisions, maintains network state, and manages connections
• user/data plane = forwards user traffic according to those decisions

The 5G Core decomposes the 4G MME into two elements:

1. Access and Mobility Management Function (AMF)
   • manages device registration, access, and mobility
   • handles signaling between the device and the core network
2. Session Management Function (SMF)
   • manages user sessions
   • communicates with the UPF to control user traffic
   • manages IP address allocation and session policies