4G LTE • LTE-ADVANCED • IMS

LTE, LTE-Advanced and IMS

Architecture • Protocols • Air Interface • Procedures • LTE-Advanced • IMS • VoLTE • SRVCC

Complete theory section: The supplied LTE/LTE-Advanced/IMS material has been kept as a full study reference here, including theory, procedures, tables, examples, comparisons, acronyms and revision material.

1. What LTE, LTE-Advanced and IMS actually do

LTE is the radio-access evolution used in the Evolved Packet System. The LTE part is primarily concerned with the radio interface between the UE and the eNodeB, while EPS includes the E-UTRAN access network and EPC core. The supplied material emphasizes that the 4G system works in packet-service mode: it transports IP packets and does not itself process traditional telephone signaling. This distinction is the foundation for understanding VoLTE.

LTE-Advanced is not a completely separate network. It is an evolution of LTE that adds techniques such as Carrier Aggregation, higher-order MIMO capability, higher modulation orders in later releases and optimized device categories. The important architectural idea is that the LTE/EPC foundation is retained while the radio interface and device capabilities become more capable.

IMS is a service-control network above IP connectivity. For VoLTE, the LTE/EPC network supplies IP transport and QoS bearers while IMS handles registration, authentication context, SIP session control, service logic, routing, supplementary services and interworking. The supplied material explicitly describes separate bearers for IMS signaling and voice media, with QCI 5 and QCI 1 respectively in its VoLTE model.

Three-layer mental model Radio layer: UE ↔ eNodeB. Core transport layer: eNodeB ↔ EPC ↔ PDN. Service layer: IMS ↔ SIP/SDP ↔ VoLTE. A good LTE/IMS answer should always explain which layer is responsible for the function being discussed.

2. EPS architecture — complete functional view

The E-UTRAN contains only one major radio-node type, the eNodeB. The EPC contains control-plane and user-plane functions. The supplied material describes the MME as the control tower, while S-GW and P-GW handle user traffic. The eNodeB performs radio-resource management, bearer allocation, mobility functions, paging handling and packet scheduling.

DomainEntityWhat it doesWhy it matters
AccessUEUser device; performs measurements, synchronization, signaling and IP trafficEndpoint of radio link
AccesseNodeBRadio resource control, scheduling, bearer mapping, mobility and radio securityCentral LTE radio node
ControlMMENAS signaling, authentication, mobility, paging, bearer control and gateway selectionControls the UE/EPC state
User planeS-GWGTP user-data forwarding and mobility anchoringCarries packets through EPC
User planeP-GWConnects EPC to PDNs; IP address allocation; policy enforcementInternet/IMS connectivity
SubscriberHSSSubscriber profile and authentication dataIdentity and security source
PolicyPCRFPolicy/QoS/charging rulesControls service-specific QoS

One of the most useful ways to understand the architecture is to separate signaling from traffic. The MME mainly processes signaling; the S-GW and P-GW primarily forward user packets. This separation makes the EPC easier to scale and helps explain why an LTE data packet can cross several GTP tunnels while control messages follow different signaling paths.

Figure: LTE/EPC architecture and interfaces.

3. eNodeB in depth

The eNodeB is much more than a radio transmitter. It schedules resources, manages radio bearers, receives UE measurements, participates in handover decisions, handles paging from the MME and performs radio-side security. The source also describes QCI-based scheduling and DSCP marking of IP packets.

3.1 eNodeB responsibilities

  • Radio resource management: decides which UE receives which time-frequency resources.
  • Bearer management: maps logical EPS bearers into radio bearers.
  • Scheduling: uses channel measurements and QoS information to choose resource assignments.
  • Mobility: uses UE measurement reports to determine when handover may be appropriate.
  • Paging: receives paging from the MME and broadcasts it within the relevant cell.
  • Security: encrypts user data and protects RRC signaling on the radio interface.
  • Cell broadcast information: provides radio configuration and cell-access information needed by UEs.
  • Traffic handling: forwards user packets toward the S-GW and from the S-GW toward the UE.

3.2 eNodeB interfaces

InterfaceEndpointsMain information
LTE-UuUE ↔ eNodeBRadio traffic, RRC signaling and transport of NAS signaling
X2eNodeB ↔ eNodeBHandover coordination, traffic and cell-load information
S1-MMEeNodeB ↔ MMES1-AP signaling, paging, mobility and bearer control
S1-UeNodeB ↔ S-GWUser-plane traffic

The source specifically states that S1-U carries mobile traffic only, while S1-MME carries signaling and NAS transport.

4. EPC nodes in depth

4.1 MME

The MME is the main control-plane entity. It does not carry ordinary user IP packets. Instead, it establishes and changes the conditions under which packets can flow. It manages EPS mobility states, NAS procedures, authentication/security, paging and bearer control. It also selects or coordinates the serving gateways.

4.2 S-GW

The S-GW is a user-plane anchor. It forwards GTP traffic between eNodeBs and the P-GW, and it can maintain continuity when the UE moves between eNodeBs. The supplied material describes S-GW anchoring within 4G and its role during inter-system packet-switched mobility. It also explains that when an idle UE has pending traffic, the S-GW informs the MME so the UE can be paged.

4.3 P-GW

The P-GW is the gateway between the EPC and external packet networks. It assigns an IPv4 or IPv6 address to the UE, applies traffic policy through its PCEF function, supports charging-related information and connects toward the PDN through SGi. The PCRF sends policy rules to the P-GW over Gx.

4.4 HSS

The HSS stores subscriber information and authentication/security data. It is consulted by the MME for LTE authentication and by IMS functions for IMS registration and service profile information.

4.5 PCRF

The PCRF translates service requirements into policy and QoS rules. In VoLTE, the P-CSCF communicates service/media requirements to the PCRF over Rx; the PCRF then interacts with the EPC policy enforcement function through Gx.

5. LTE interfaces and reference points

Reference pointPurposeProtocol family / data
LTE-UuUE ↔ eNodeBRRC, PDCP, RLC, MAC, physical layer
X2eNodeB ↔ eNodeBHandover/cell information and user-plane support
S1-MMEeNodeB ↔ MMES1-AP; NAS transported inside S1-AP/RRC
S1-UeNodeB ↔ S-GWGTP user plane
S11MME ↔ S-GWGTP-C control
S5/S8S-GW ↔ P-GWGTP control/user plane
S6aMME ↔ HSSDiameter subscriber/authentication signaling
GxP-GW/PCEF ↔ PCRFPolicy/QoS rules
SGiP-GW ↔ PDNIP traffic
RxP-CSCF ↔ PCRFIMS service/media QoS requirements

Understanding interfaces is essential because LTE procedures are distributed across them. For example, default-bearer establishment uses GTP-C on S11/S5, S1-AP on S1-MME, and RRC reconfiguration on LTE-Uu.

6. LTE protocol architecture

Figure: simplified LTE protocol-stack view.

6.1 NAS

NAS is the UE-to-MME control layer. It is not a radio protocol in the same sense as RRC; it is transported through the radio and access network to reach the MME. NAS procedures include EPS mobility management (EMM) and EPS session management (ESM). Attach, authentication, security mode, tracking-area update and service request are major EMM procedures.

6.2 RRC

RRC is the UE-to-eNodeB control protocol. It establishes and releases RRC connections, carries system information, controls radio bearers, supports security activation, measurement reporting and handover commands. NAS messages are transported through RRC.

6.3 PDCP/RLC/MAC

  • PDCP performs packet processing and radio-interface security functions.
  • RLC provides segmentation/reassembly and acknowledged or unacknowledged transfer depending on the bearer.
  • MAC multiplexes logical channels, supports scheduling-related functions and handles HARQ operation.
  • The physical layer converts scheduled transport blocks into actual radio transmissions.

For VoLTE, the source specifically emphasizes the use of an unacknowledged data radio bearer for the voice stream because retransmitting a lost voice packet can make it arrive too late. Other signaling and default-bearer traffic can remain on acknowledged bearers.

7. LTE air interface fundamentals

LTE divides radio resources in both time and frequency. The frequency axis is divided into orthogonal subcarriers and the time axis into OFDM symbols, slots and subframes. This two-dimensional resource structure is what allows the eNodeB scheduler to give different users different resource blocks.

7.1 Downlink OFDMA

OFDMA uses many narrow, mutually orthogonal subcarriers. A high-rate data stream is divided across parallel lower-rate streams. Orthogonality lets adjacent subcarriers be closely spaced without the same type of inter-carrier interference that would occur if they were independent non-orthogonal carriers.

7.2 Uplink SC-FDMA

LTE uses SC-FDMA in the uplink to reduce peak-to-average power variation compared with ordinary OFDMA. This matters because the UE has a battery-powered power amplifier, and a lower peak-to-average requirement can improve transmitter efficiency and practical battery behavior.

7.3 Cyclic prefix

Multipath propagation creates delayed copies of the transmitted signal. The cyclic prefix adds a guard interval so delayed energy from one symbol is less likely to destroy the orthogonality of the next symbol. LTE defines normal and extended cyclic-prefix configurations. The supplied material gives 15 kHz subcarrier spacing as the normal basis.

8. OFDM, cyclic prefix and SC-FDMA theory

An OFDM symbol can be understood as the sum of many orthogonal sinusoidal subcarriers. Each subcarrier carries a complex modulation symbol such as QPSK or QAM. The receiver uses an FFT-based process to separate the subcarriers. Because the subcarriers are orthogonal over the useful symbol interval, they can overlap in frequency while remaining separable at the receiver.

The cyclic prefix does not carry new user information; it is a guard mechanism. Its trade-off is that a longer prefix improves multipath tolerance but consumes more of the available symbol time. The source describes a normal prefix of approximately 4.7 microseconds and an extended configuration for larger delay spreads.

SC-FDMA adds a transform-spreading step before the OFDM modulation process. The practical consequence is a more single-carrier-like waveform with lower peak-to-average power variation, which is advantageous for the mobile transmitter.

9. LTE resource grid and physical channels

Figure: LTE resource-grid and timing concept.

9.1 Resource Block

The Resource Block is the fundamental time-frequency unit assigned to a UE. The supplied source describes an RB as 0.5 ms on the time axis and 180 kHz on the frequency axis. With a normal cyclic prefix it contains 12 subcarriers × 7 OFDM symbols = 84 resource elements. With an extended cyclic prefix it contains 72 resource elements. The number of RBs depends on carrier bandwidth: 6, 15, 25, 50, 75 and 100 for 1.4, 3, 5, 10, 15 and 20 MHz respectively.

BandwidthUsable bandwidth in source tableNumber of RBs
1.4 MHz1.08 MHz6
3 MHz2.7 MHz15
5 MHz4.5 MHz25
10 MHz9 MHz50
15 MHz13.5 MHz75
20 MHz18 MHz100

9.2 Downlink signals/channels

ElementPurpose
PSSPrimary synchronization and part of physical-cell identification
SSSFrame synchronization, cell identification, FDD/TDD and cyclic-prefix-related information
Reference SignalsRadio-channel estimation and synchronization support
PBCHCell broadcast/system information transport
PDCCHDownlink control: resource allocation and uplink resource assignments
PCFICHIndicates the number of OFDM symbols used by PDCCH
PHICHHARQ acknowledgment for uplink data
PDSCHDownlink unicast traffic and signaling
PMCHMulticast traffic

These functions and channel roles are summarized directly from the supplied LTE-interface material.

9.3 Uplink signals/channels

ElementPurpose
PRACHRandom access
PUCCHUplink control information such as acknowledgments, channel quality and scheduling requests
PUSCHUnicast uplink traffic
SRSSounding/reference information used to assess the uplink radio channel
DRS/DM-RSReference signal used for demodulation/channel estimation

The source describes SRS as using Zadoff-Chu sequences and notes that PUCCH carries UCI when no PUSCH resource is available.

10. MIMO, diversity, precoding and beamforming

MIMO uses multiple transmit and receive antennas to exploit the spatial characteristics of the radio channel. If the channel provides sufficiently independent paths, multiple data streams can be transmitted simultaneously over the same time-frequency resources. This increases throughput without requiring additional spectrum.

10.1 Diversity vs spatial multiplexing

Diversity sends related information through different spatial paths so that a deep fade affecting one path is less likely to destroy the whole transmission. Spatial multiplexing instead sends independent streams, using the channel's spatial degrees of freedom to increase data rate.

10.2 Beamforming

Beamforming changes the relative phases/amplitudes of signals transmitted by multiple antennas so that the combined signal is stronger in a desired direction. The source describes MISO beamforming as a way of shaping the beam and improving SNR or range.

10.3 LTE transmission modes

  • SIMO: multiple receive paths with one transmit path.
  • MISO: multiple transmit antennas with a single receive antenna; can provide diversity/beamforming benefits.
  • SU-MIMO: multiple spatial streams for one UE.
  • MU-MIMO: spatial resources are used for multiple UEs, increasing cell capacity.
  • Precoding: transforms the data before transmission so the antenna signals combine in a desired way.

11. LTE scheduling, link adaptation and HARQ

LTE scheduling is performed at short time intervals. The source describes subframe-level scheduling, meaning the eNodeB can reconsider resource allocation roughly every millisecond. The scheduler uses information about channel conditions, available data and QoS. This creates a trade-off between maximizing instantaneous throughput and maintaining fairness among users.

11.1 Link adaptation

Link adaptation changes the modulation and coding scheme according to the radio conditions. Under poor conditions, a robust lower-order modulation such as QPSK can be used; under better conditions, 16-QAM or 64-QAM can carry more bits per symbol. The selected modulation must be supported by the measured channel quality.

11.2 HARQ

HARQ combines forward error correction with fast retransmission. If the receiver cannot decode a transport block, it sends negative feedback and the transmitter can send additional information. The goal is to recover from errors without the large delay associated with higher-layer retransmission alone.

11.3 Semi-persistent scheduling

For predictable periodic traffic such as voice, semi-persistent scheduling can reduce the overhead of assigning resources dynamically for every transmission. The supplied VoLTE material describes this as one way to maintain resources for the voice bearer.

12. EPS bearers and QoS

Figure: default and dedicated EPS bearers.

A bearer is a logical transport path with a defined QoS treatment. The default bearer provides the UE's basic PDN connectivity. A dedicated bearer is added when a specific traffic flow needs different QoS characteristics. The source explicitly describes the dedicated bearer as being associated with a QCI different from the default bearer.

12.1 QCI

QCI identifies standardized QoS characteristics used to determine how traffic should be treated. In the supplied VoLTE model, QCI 5 is used for IMS signaling and QCI 1 for the voice media bearer. The later LTE source explains that QCI 1 has tighter delay handling than the example QCI 9 used for general Internet traffic.

12.2 Why voice needs a dedicated bearer

Voice is interactive and time-sensitive. A lost voice packet is often less useful after a long retransmission delay than a lost web packet. Therefore, the voice bearer is designed to prioritize timely delivery and tolerate a controlled amount of packet loss. The supplied source describes a guaranteed bit rate example of 40 kbit/s and traffic filters that identify the RTP voice flow using IP address and UDP-port information.

13. LTE attach, authentication and security

Attach is the procedure through which the UE becomes registered with the EPS and obtains the necessary connectivity and bearer context. It involves NAS signaling between UE and MME, authentication using HSS information, security activation and default-bearer establishment.

13.1 Authentication theory

The MME obtains authentication material from the HSS and challenges the UE. The UE produces a response value. The MME compares the response with the expected value. If they match, authentication succeeds and the MME starts security mode. The source shows EMM AUTHENTICATION RESPONSE followed by EMM SECURITY MODE COMMAND and EMM SECURITY MODE COMPLETE. After successful security activation, subsequent NAS signaling is encrypted and integrity checked according to the selected algorithms.

13.2 Two security domains

  • NAS security protects the UE ↔ MME signaling relationship.
  • RRC/user-plane security protects traffic across the LTE-Uu radio interface between UE and eNodeB.
  • The supplied LTE evolution material explains that integrity protection is mandatory for signaling and that ciphering can be activated for signaling/user traffic according to the selected algorithms.

13.3 Why old GSM-only authentication data is insufficient

The LTE source explains that a GSM-only SIM profile may lack the LTE authentication information expected by the HSS. In that case, the MME cannot complete LTE authentication and the device may move toward UMTS/GSM. This illustrates an important principle: LTE access depends not only on radio capability but also on subscriber authentication data in the core network.

14. Tracking Area Update, paging and Service Request

14.1 Tracking Area Update

A UE in registered state periodically updates its tracking information and also performs an update when its Tracking Area Code changes. It sends EMM TRACKING AREA UPDATE REQUEST to the MME. The MME returns ACCEPT or REJECT. If a new GUTI is assigned, the UE confirms it with TRACKING AREA UPDATE COMPLETE.

14.2 Paging

When a UE is idle, it is not maintaining an active radio connection to an eNodeB, but it remains registered in the core. If downlink data or a service is waiting, the network pages the UE. The eNodeB broadcasts the paging information within the cell so the UE can return to connected operation.

14.3 Service Request

The UE responds to pending signaling/traffic by sending EMM SERVICE REQUEST. The procedure re-establishes the required LTE-Uu and S1-U bearers. The source states that paging informs the UE that traffic is waiting and Service Request is then used to restore the active bearer path.

15. PDN connection and default bearer establishment

The default bearer gives the UE its basic packet-data connectivity. It is not just one message; several interfaces participate. The supplied source explicitly lists GTP-C CREATE SESSION REQUEST/RESPONSE on S11/S5, S1-AP INITIAL CONTEXT SETUP REQUEST/RESPONSE on S1-MME and RRCConnectionReconfiguration / Complete on LTE-Uu.

  1. UE requests/indicates the need for PDN connectivity as part of the EPS procedure.
  2. MME selects the appropriate S-GW/P-GW path and requests session creation.
  3. P-GW allocates IP connectivity and applies policy.
  4. S-GW/P-GW create the GTP tunnel context.
  5. MME asks the eNodeB to establish the required radio context.
  6. eNodeB performs RRC reconfiguration and establishes the radio bearer.
  7. UE confirms the configuration and completes the attach/default-bearer procedure.

The key theory point is that a single user IP flow is represented by a chain of logical contexts: radio bearer on LTE-Uu, S1-U tunnel, S5/S8 tunnel and PDN-side IP connectivity.

16. LTE mobility and inter-RAT operation

Mobility exists at several levels. Intra-LTE handover moves the UE from one eNodeB/cell to another. Inter-RAT mobility moves the UE from LTE to UMTS or GSM. The network tries to make the transition efficient by informing the UE about neighboring frequencies and radio technologies rather than forcing it to scan everything blindly.

16.1 LTE-to-LTE handover

  • UE measures neighboring cells and reports measurement information.
  • Serving eNodeB decides that a handover should be initiated.
  • X2 or S1 procedures coordinate target-cell preparation.
  • Target resources and security/context information are prepared.
  • UE receives the handover command and synchronizes to the target cell.
  • User traffic is forwarded through the prepared path and the old context is released.

16.2 LTE-to-UMTS/GSM

The supplied LTE evolution material describes three basic procedures: cell reselection, RRC connection release with redirection, and inter-RAT handover from LTE to UMTS. These mechanisms reduce the time required to find another network layer when LTE coverage becomes insufficient.

17. CSFB and its limitations

CSFB is a transitional voice solution. Because baseline LTE is packet-switched, the UE can be moved from LTE to a legacy 2G/3G circuit-switched system to place or receive a telephone call. The SGs interface connects the MME with the MSC Server for CSFB-related signaling. The source describes SGs-AP functions for mobile attachment, location update and paging.

17.1 Outgoing CSFB concept

  1. UE is attached to LTE and indicates CSFB capability.
  2. MME has an SGs relationship with the MSC Server.
  3. When a voice call is requested, LTE coordinates the transition to the target 2G/3G cell.
  4. UE sends CM SERVICE REQUEST to the MSC after moving to the circuit-switched access network.
  5. When the call ends, the UE can return to LTE and reactivate suspended packet bearers.

17.2 Incoming CSFB concept

For an incoming call, the MSC Server can page the UE through the MME using SGs-AP. The MME and eNodeBs then page the UE over LTE. The UE responds with an Extended Service Request and the network completes the move toward the circuit-switched voice domain.

Limitation CSFB interrupts the LTE packet-service path for voice. VoLTE avoids this by keeping voice on LTE and using IMS for telephone-service signaling.

18. LTE-Advanced — why it was required

One LTE carrier has a finite bandwidth. Once a network has many users, simply improving the modulation or adding more antennas cannot always compensate for fragmented spectrum and growing traffic. LTE-Advanced therefore combines several enhancements: more spectrum through Carrier Aggregation, improved spatial processing, higher modulation orders in later releases and new device categories.

18.1 Carrier Aggregation

Carrier Aggregation allows several LTE component carriers to be used together. The source explains that CA initially helped combine fragmented spectrum—for example, two 10 MHz carriers in different bands—to create an aggregate 20 MHz resource. Today its value is both peak-rate improvement and, importantly, better dynamic scheduling across spectrum with different propagation characteristics.

18.2 Why CA is not simply 'double speed'

Aggregate bandwidth may increase, but a real user only benefits when the device supports the band combination, radio conditions permit simultaneous use and the scheduler has traffic to send. The source emphasizes that as more users share the cell, average speeds are lower than theoretical peaks.

19. Carrier Aggregation in depth

19.1 Component carriers

  • PCell/PCC: primary cell/component carrier that anchors the connection.
  • SCell/SCC: secondary carriers added to increase available bandwidth.
  • Configuration is slower and signaling-heavy; activation/deactivation is faster and uses MAC-layer control.
  • Cross-carrier scheduling can allow assignments for several component carriers to be announced on the PCell PDCCH.

The source explains that cross-carrier scheduling can reduce the need for the UE to monitor every component carrier's PDCCH and that SCells can be activated/deactivated depending on traffic load. This saves UE power when extra carriers are not needed.

19.2 Four CA types

TypeTheoryExample
Inter-bandCarriers in different frequency bands800 + 1800 MHz
Intra-band contiguousAdjacent carriers in one band20 + 10 MHz contiguous spectrum
Intra-band non-contiguousSeparated blocks within one bandTwo separated chunks in the same band
Inter-band FDD/TDDCombines FDD and TDD carriers in supported configurationsFDD DL + TDD DL

The source identifies inter-band CA as common in practice and explains the hardware complexity on the UE side, where extra RF hardware, antennas and processing may be required.

19.3 Uplink CA limitations

Uplink CA is more constrained than downlink CA because the UE has limited transmit power and limited RF hardware. The source notes that uplink CA is often limited to two carriers and that cell-edge power requirements can make simultaneous uplink transmission impractical. It also notes hardware limitations when different bands share transmitter chains.

20. LTE-Advanced device categories and throughput

CategoryExample maximum DL with CA (Mbit/s)Typical DL carriersMax UL (Mbit/s)Receive antennas
Cat 41501502
Cat 63002502
Cat 94503502
Cat 1260041002
Cat 18120051254

These figures are source-table examples for device categories, not guaranteed field performance. The source also notes practical peak speeds around 200–250 Mbit/s per 20 MHz carrier under ideal conditions with 4×4 MIMO, while interference, multiple users and reception conditions reduce actual throughput.

21. 256-QAM, higher MIMO and practical throughput

256-QAM maps 8 bits to each modulation symbol. It can increase throughput because more bits are transmitted for the same number of symbols. However, the higher constellation points are closer together, so the receiver needs a cleaner radio channel. The supplied material notes that 256-QAM is generally used only in very good radio conditions and over a relatively small area near a macro-cell site.

Similarly, higher-order MIMO increases potential throughput only if the propagation channel provides enough independent spatial paths and the network and UE both support the configuration. Therefore, a device specification such as '4×4 MIMO' is a capability, not a promise that every location will operate with four simultaneous streams.

22. LTE IoT evolution: Cat-0, Cat-M1 and NB-IoT

LTE was originally designed for broadband devices, but IoT devices often have very different requirements: low cost, low power, small amounts of data and deep indoor coverage. The supplied material describes LTE Cat-0, Cat-M1 and NB-IoT as adaptations that reuse existing LTE infrastructure.

22.1 Cat-0

Cat-0 reduces device complexity by limiting throughput to around 1 Mbit/s. Optional half-duplex operation can reduce RF complexity and power consumption. Power Save Mode allows the device to stop monitoring paging for long periods while retaining its registration state.

22.2 Cat-M1

Cat-M1 targets low-power IoT while retaining a more flexible LTE-like operation than NB-IoT. It uses a narrower device bandwidth and aims at roughly 1 Mbit/s class operation in the supplied material, with power/complexity improvements.

22.3 NB-IoT

NB-IoT is optimized further for very low-cost, low-power devices, small and infrequent data transfers and improved deep indoor coverage. The supplied material describes a 180 kHz channel and LTE-derived protocol/security concepts while intentionally reducing functions such as connected-mode handover and measurement/reporting complexity.

The important engineering point is that these IoT categories do not require a completely new radio network. The source states that existing LTE infrastructure can support traditional broadband and the newer MTC/CIoT device categories simultaneously, with software enhancements to eNodeBs/core components.

23. IMS — purpose and architecture

The IMS network provides telephone service when the mobile network operates in packet-service mode. It connects to the EPC P-GW, PCRF, HSS and external PSTN/PLMN networks. The supplied material lists CSCF functions, Application Servers, Multimedia Resource Functions, interconnection gateways and charging functions as major IMS components.

Figure: IMS architecture for VoLTE.

23.1 Why IMS is separate from LTE

LTE knows how to transport IP packets efficiently, but it does not by itself define the complete telephone-service logic. IMS supplies that missing service layer. It knows user identities, registration state, session routing, supplementary services and interworking rules. This separation allows the same packet transport network to carry voice, video and other IP services.

24. P-CSCF, I-CSCF, S-CSCF and supporting functions

24.1 P-CSCF

The P-CSCF is the UE's first IMS contact and acts primarily as a SIP proxy. It forwards REGISTER and INVITE messages, detects emergency calls, participates in charging information generation, establishes the security relationship with the UE and communicates with the PCRF to request the resources required by the session.

24.2 I-CSCF

The I-CSCF is an IMS entry/routing point. During initial registration it communicates with the HSS and helps select the S-CSCF. For incoming calls, it can query the HSS to locate the S-CSCF responsible for the destination subscriber.

24.3 S-CSCF

The S-CSCF is the central session-control function. It acts as registrar during registration, a location server by maintaining the relationship between user identity and IP address, and a SIP proxy during session establishment. It authenticates the UE using HSS information, downloads the subscriber profile and routes calls toward Application Servers, other IMS networks, P-CSCFs or gateways.

24.4 E-CSCF

The E-CSCF handles emergency calls. It can use the Location Retrieval Function to obtain or validate UE location information and route the emergency request to the appropriate emergency center.

24.5 Application Servers

Application Servers extend SIP/IMS behavior. In VoLTE, an MMTel Application Server provides supplementary telephone services such as call forwarding, conference calls, call hold and caller-identity related services.

25. IMS registration and authentication

Figure: IMS registration and authentication.

When a VoLTE device powers on, it establishes IP connectivity and then registers with IMS using SIP. The source describes a separate LTE default bearer for IMS services rather than using the ordinary Internet bearer for IMS signaling. During registration, the UE identifies the IMS network and advertises supported authentication/ciphering capabilities and supported services.

  1. UE sends SIP REGISTER toward the P-CSCF.
  2. P-CSCF adds Path information and forwards the request.
  3. I-CSCF queries HSS and helps identify the S-CSCF.
  4. S-CSCF obtains authentication information and returns 401 Unauthorized with a challenge.
  5. UE/SIM calculates the authentication response and security parameters.
  6. UE sends the second REGISTER with the authentication response.
  7. S-CSCF verifies the response and retrieves the IMS subscriber profile from HSS.
  8. S-CSCF returns 200 OK and Service-Route information.
  9. UE can subscribe to registration-state notifications.

The source explains that IMS signaling security is based on the subscriber secret key and authentication algorithms, with integrity protection mandatory and encryption optionally used for the UE-to-P-CSCF SIP signaling relationship. After successful authentication, the S-CSCF downloads the subscriber's IMS profile.

26. SIP and SDP theory for VoLTE

SIP is the signaling protocol used to create, modify and terminate IMS sessions. It does not carry the voice samples themselves. Instead, SIP establishes the session and negotiates the media parameters. The actual speech packets are normally carried using RTP over UDP/IP.

26.1 Important SIP messages

MessagePurpose in VoLTE
REGISTERRegister UE with IMS
INVITERequest session establishment
100 TryingConfirms that the request is being processed
183 Session ProgressProvides early session/media information
180 RingingDestination is alerting
200 OKSuccessful response / call establishment
ACKConfirms final INVITE response
BYETerminates established session
401 UnauthorizedAuthentication challenge during registration

26.2 SDP

SDP is embedded in SIP messages and describes media. For a voice call it can contain supported codecs, IP addresses and RTP ports. The P-CSCF uses this information as part of the QoS-resource process. The source explicitly states that SDP information in INVITE/Session Progress is needed to select a compatible speech codec and establish the dedicated bearer.

27. Detailed VoLTE call establishment

Figure: detailed VoLTE call establishment.
  1. Originating UE sends SIP INVITE containing SDP toward the P-CSCF.
  2. P-CSCF responds with 100 Trying and forwards the INVITE to the S-CSCF.
  3. Originating S-CSCF checks service requirements and routes toward the destination IMS domain.
  4. I-CSCF identifies the destination subscriber's S-CSCF using HSS information.
  5. Destination S-CSCF determines the destination P-CSCF and forwards the INVITE.
  6. Destination P-CSCF delivers the INVITE to the destination UE.
  7. Destination UE returns SDP/media information in a provisional response.
  8. P-CSCF/PCRF interaction requests the resources required for the voice bearer.
  9. LTE establishes a dedicated voice bearer with the required QoS.
  10. Destination UE alerts the user and sends 180 Ringing.
  11. 200 OK and ACK complete SIP session establishment.
  12. RTP voice packets then use the dedicated voice bearer while SIP remains the control layer.

The supplied LTE-Advanced source describes this sequence in detail: P-CSCF sends 100 Trying, S-CSCF routes through I-CSCF, SDP is used for codec/resource negotiation, the P-CSCF works with PCRF to establish the dedicated bearer, and speech packets can then flow directly through the packet network.

28. VoLTE dedicated bearer and QoS

A VoLTE call normally uses more than one bearer. IMS signaling has its own QoS treatment, while the voice RTP flow receives a dedicated bearer. The supplied source explicitly identifies QCI 5 for signaling and QCI 1 for voice.

28.1 Why RLC unacknowledged mode is used for voice

Suppose a voice packet is lost. A retransmission that arrives too late cannot reconstruct the already-played speech. Therefore the voice bearer can use RLC unacknowledged mode. The source states that the voice bearer is configured to tolerate controlled packet loss and that radio parameters are chosen so packet loss stays within a range the codec can tolerate.

28.2 QCI 1 example

The supplied material gives a QCI 1 example with a 100 ms maximum queueing-delay target under congestion versus 300 ms for the QCI 9 example used for Internet traffic. It also describes a 40 kbit/s guaranteed rate example and traffic filters that restrict the dedicated bearer to the voice flow. These values should be treated as the source's example configuration, not as a universal guarantee for every operator.

29. IMS supplementary services and interworking

29.1 Supplementary services

  • Call forwarding
  • Call hold
  • Conference calling
  • Calling-party identity presentation/restriction
  • Other MMTel supplementary services implemented through Application Server logic

29.2 PSTN/PLMN interworking

When the destination is not an IMS subscriber, IMS needs gateway functions to convert between SIP/IP multimedia and circuit-switched signaling/media. The supplied source lists BGCF, MGCF, MGW and SGW/Signaling Gateway functions for this interconnection.

30. CSFB vs VoLTE vs SRVCC

FeatureCSFBVoLTESRVCC
Where voice is carried2G/3G circuit-switched networkLTE packet networkStarts on LTE, transfers to 2G/3G CS
IMS requiredNoYesYes
LTE maintained during callNo, UE moves to legacy RATYesNo after transfer
QoS mechanismLegacy CS resource reservationDedicated LTE EPS bearerLTE bearer before transfer + CS bearer after transfer
Main purposeEarly LTE voice solutionNative LTE voice serviceVoice continuity when LTE coverage is lost
Typical signalingCS/SGs mechanismsSIP/SDPIMS + MME/MSC transfer signaling

CSFB is a fallback mechanism; VoLTE is the normal IP-voice service; SRVCC is a continuity mechanism for an already established IMS voice call when LTE coverage becomes insufficient.

31. SRVCC architecture and detailed procedure

Figure: VoLTE media/signaling and SRVCC continuity.

SRVCC allows an ongoing VoLTE call to move from LTE packet service to a 2G/3G circuit-switched call while preserving the communication. The key reason for the ATCF/ATGW architecture is that the voice path is anchored so it can be changed quickly during the transfer.

31.1 SRVCC functions

EntityRole
SCC ASControls service continuity and stores/uses information required for transfer
ATCFAnchors SIP signaling and coordinates the access transfer
ATGWAnchors RTP media and can perform codec transcoding
MMECoordinates the packet-to-circuit transfer with the MSC Server
MSC Server/MGWReserves and provides the target circuit-switched resources

The source describes the SCC AS as controlling SRVCC, ATCF as the SIP signaling anchor and ATGW as the RTP anchor. The ATGW can also transcode voice when LTE and GSM/UMTS codecs differ.

31.2 SRVCC handover sequence

  1. UE is in an active VoLTE call.
  2. UE sends measurement reports; eNodeB determines that LTE-to-CS handover is required.
  3. eNodeB sends S1-AP HANDOVER REQUIRED to the MME.
  4. MME identifies the voice flow using QCI 1 and separates it from other packet flows for relocation.
  5. MME sends a transfer request toward the MSC Server, including the SRVCC session information.
  6. MSC Server reserves the required circuit-switched resources in the target UMTS/GSM cell.
  7. MME requests the ATCF to prepare the ATGW for the media transfer.
  8. ATCF/ATGW prepare the new RTP endpoint toward the MSC media gateway.
  9. MSC Server confirms preparation; MME sends the handover command to the UE.
  10. UE moves to the target GSM/UMTS cell and continues voice over the circuit-switched bearer.
  11. MSC media gateway converts the circuit-switched media toward the RTP-anchored path.
  12. Other packet services can be handled according to the mobility procedure and available target-network support.

The source gives this sequence in detail, including MME PS-to-CS transfer signaling, MSC resource reservation, ATGW preparation and the final handover command.

31.3 Why the RTP anchor is important

Without an anchor, the endpoints of the RTP session would need to be changed abruptly during the handover. The ATGW provides a stable media point while one side of the call changes from an IP endpoint to an MSC media gateway. This makes the transition controllable and also provides a place for transcoding if required.

32. NFV and evolution of LTE network implementation

The LTE evolution material describes a shift from dedicated telecom hardware toward standardized computing and networking platforms. The separation of signaling-heavy functions such as MME from packet-forwarding functions such as S-GW makes virtualization practical. Later, network functions can run inside virtual machines or other software-defined environments.

32.1 Why virtualization matters

  • Capacity can be increased by adding software instances rather than installing a fixed proprietary appliance.
  • Signaling and database functions can use standard compute/storage platforms.
  • Network functions become less dependent on a specific hardware platform.
  • Operators can scale resources more flexibly according to traffic demand.

This topic belongs to LTE-Advanced Pro's broader evolution because the improvement is not only radio speed; the network implementation itself becomes more software-oriented.

33. End-to-end example: mobile powers on → VoLTE call

  1. UE powers on and searches for LTE synchronization and cell information.
  2. UE establishes radio access and performs random access/RRC procedures.
  3. UE sends NAS attach signaling through the eNodeB to the MME.
  4. MME authenticates the UE using HSS information.
  5. NAS and radio security are activated.
  6. MME establishes the default EPS bearer through S-GW/P-GW and the eNodeB.
  7. UE obtains IP connectivity.
  8. UE establishes the IMS bearer and sends SIP REGISTER.
  9. I-CSCF/S-CSCF/HSS perform IMS registration and authentication.
  10. S-CSCF downloads the IMS profile and returns 200 OK.
  11. User initiates a call; UE sends SIP INVITE with SDP.
  12. P-CSCF/S-CSCF/I-CSCF route the INVITE to the destination.
  13. PCRF receives the service/QoS requirement and causes the EPC to establish a dedicated voice bearer.
  14. Destination answers; SIP 200 OK and ACK complete session establishment.
  15. RTP voice packets flow over the dedicated bearer.
  16. If LTE coverage becomes insufficient, eNodeB/MME/MSC/ATCF/ATGW can prepare SRVCC.
  17. UE moves to 2G/3G and the voice call continues through the circuit-switched path.
End-to-end takeaway Every stage has a different owner: eNodeB controls radio resources, MME controls mobility/signaling, EPC gateways provide IP transport, HSS provides subscriber/authentication data, IMS controls the voice session, PCRF controls QoS policy, and SRVCC provides continuity when LTE coverage ends.

34. Detailed comparison tables

TopicLTELTE-AdvancedIMS/VoLTE
Primary role4G radio + packet connectivityHigher capacity/throughput evolution of LTEService/session control over IP
Main nodeseNodeB + EPCSame foundation + enhanced radio/device supportCSCFs, HSS, AS, PCRF interaction, gateways
Main protocolsRRC, NAS, S1-AP, GTP, PDCP/RLC/MACSame LTE protocols plus enhanced radio proceduresSIP, SDP, Diameter
RadioOFDMA DL / SC-FDMA ULCA, higher MIMO, later higher modulationUses LTE radio as transport
SpectrumSingle LTE carrier up to 20 MHz baselineMultiple component carriersNot a spectrum technology
VoiceCSFB or IMSVoLTE + SRVCC supported in evolutionSIP-controlled IP voice
QoSEPS bearers/QCIMore capacity for QoS trafficP-CSCF/PCRF requests service-specific QoS

34.1 Protocol-to-function map

NeedProtocol/functionMain node
Authenticate LTE subscriberNAS + HSS/MMEMME/HSS
Protect LTE signalingNAS/RRC securityMME/eNodeB
Schedule radio resourcesMAC schedulereNodeB
Transport user IPGTP-UeNodeB/S-GW/P-GW
Allocate IP addressP-GWP-GW
Apply policy/QoSPCRF/PCEFPCRF/P-GW
Register for VoLTESIP REGISTERP-CSCF/I-CSCF/S-CSCF
Authenticate IMSSIP challenge + HSSS-CSCF/HSS
Set up callSIP INVITE + SDPCSCFs
Prioritize voice mediaDedicated bearer/QCI 1PCRF/EPC/eNodeB
Move voice to 2G/3GSRVCCMME/MSC/ATCF/ATGW/SCC AS

35. Important acronyms

AcronymMeaning
AMRAdaptive Multi-Rate
ASApplication Server
ATCFAccess Transfer Control Function
ATGWAccess Transfer Gateway
BGCFBreakout Gateway Control Function
CACarrier Aggregation
CSFBCircuit Switched Fallback
CSCFCall Session Control Function
DRBData Radio Bearer
E-CSCFEmergency CSCF
EPCEvolved Packet Core
EPSEvolved Packet System
E-UTRANEvolved Universal Terrestrial Radio Access Network
FDDFrequency Division Duplex
GTPGPRS Tunneling Protocol
HSSHome Subscriber Server
I-CSCFInterrogating CSCF
IMSIP Multimedia Subsystem
LTELong Term Evolution
MACMedium Access Control
MMEMobility Management Entity
MIMOMultiple Input Multiple Output
NASNon-Access Stratum
NB-IoTNarrowband Internet of Things
OFDMAOrthogonal Frequency Division Multiple Access
P-CSCFProxy CSCF
PCRFPolicy and Charging Rules Function
P-GWPDN Gateway
PDCPPacket Data Convergence Protocol
PDSCHPhysical Downlink Shared Channel
PDCCHPhysical Downlink Control Channel
PRACHPhysical Random Access Channel
PUCCHPhysical Uplink Control Channel
PUSCHPhysical Uplink Shared Channel
QAMQuadrature Amplitude Modulation
QCIQoS Class Identifier
RLCRadio Link Control
RRCRadio Resource Control
RTPReal-Time Transport Protocol
S-CSCFServing CSCF
S-GWServing Gateway
SC-FDMASingle Carrier Frequency Division Multiple Access
SDPSession Description Protocol
SIPSession Initiation Protocol
SRVCCSingle Radio Voice Call Continuity
TDDTime Division Duplex
UEUser Equipment
VoLTEVoice over LTE

Continue learning

Explore 5G / NR, 6G, SIP and Log Analysis from the same Telecom Knowledge Hub.

Back to Knowledge Hub