Tutorial 8: LTE Architecture and Operations

Table of Contents

Learning Objectives

Upon completion of this extended tutorial, students will be able to:

Overview

Long Term Evolution (LTE) is the 4th generation mobile network standard developed by 3GPP, delivering high‑speed broadband services with low latency. This tutorial provides a detailed examination of the LTE architecture and its operational procedures. We begin with an overview of the System Architecture Evolution (SAE) and the split between E‑UTRAN and EPC. We then dissect the protocol stack—from the physical layer up to the PDCP and RRC—and trace the flow of user data. We cover mobility management (idle and connected states), handover procedures (X2 and S1), and the security mechanisms (AKA, ciphering, integrity). We also discuss QoS and bearer management, and the enhancements introduced in LTE‑Advanced (carrier aggregation, enhanced MIMO, CoMP). The tutorial concludes with case studies on VoLTE and real‑world deployment. This extended version includes rigorous technical depth and a substantial set of assessment questions.

1. Technical & Theoretical Content

8.1 Overview of LTE and SAE

LTE was introduced in 3GPP Release 8 as a major evolution from UMTS. The System Architecture Evolution (SAE) defines the new all‑IP core network (EPC). Key goals:

8.2 LTE Network Architecture: E‑UTRAN and EPC

The LTE network is divided into the Radio Access Network (E‑UTRAN) and the Core Network (EPC).

The S1 interface has two variants: S1‑MME (control plane) and S1‑U (user plane).

8.3 LTE Protocol Stack (User Plane and Control Plane)

8.3.1 User Plane

User data flows through:

8.3.2 Control Plane

Additionally includes:

8.4 LTE Data Delivery Process (Downlink and Uplink)

Downlink (eNodeB to UE):

  1. Higher‑layer packets are received by PDCP (header compression, ciphering).
  2. RLC segments into RLC PDUs (if needed) and adds sequence numbers.
  3. MAC multiplexes logical channels, performs scheduling (dynamic allocation of resource blocks), and adds MAC headers (including HARQ).
  4. PHY modulates and transmits over the air using OFDMA.

Uplink (UE to eNodeB):

  1. UE PDCP processes data, RLC segments, MAC adds headers and scheduling request (SR) for uplink grants.
  2. UE transmits using SC‑FDMA on assigned resource blocks.
  3. eNodeB MAC receives, performs HARQ combining, and forwards to RLC/PDCP for reassembly.

HARQ (Hybrid ARQ) with Chase Combining or Incremental Redundancy is used for error correction at the MAC layer.

8.5 Mobility Management: Idle and Connected Modes

The UE can be in two states:

Transition from IDLE to CONNECTED involves:

  1. Random access procedure (RACH).
  2. RRC connection setup.
  3. Service request and bearer establishment.

8.6 Handover Procedures (X2 and S1 Handover)

X2‑based handover: When UE moves between eNodeBs that have an X2 interface (direct link), handover can be performed with minimal involvement of the core network. Steps:

  1. UE measurement report triggers handover decision.
  2. Source eNodeB sends Handover Request to target eNodeB over X2.
  3. Target eNodeB performs admission control and replies with Handover Request Acknowledgment.
  4. Source eNodeB sends RRC Connection Reconfiguration to UE (including target cell info).
  5. UE synchronises to target and completes handover.
  6. Target eNodeB sends Path Switch Request to MME to update SGW/PGW.

S1‑based handover: Used when X2 is not available (e.g., different MME pools). Involves MME and SGW relocation, with more signalling.

8.7 LTE Security: AKA, Encryption, and Integrity

LTE security is built on the EPS AKA (Authentication and Key Agreement) procedure:

Encryption (ciphering) and integrity protection are applied to user plane and control plane traffic using AES (128‑bit) or SNOW 3G (for legacy).

8.8 Quality of Service (QoS) and Bearer Management

LTE uses the concept of EPS bearers to provide QoS. A bearer is a logical path with a specific set of QoS parameters (QCI, ARP, GBR vs. Non‑GBR).

Bearer establishment is initiated by the UE (UL) or network (DL) via NAS messages (Bearer Resource Allocation Request).

8.9 LTE‑Advanced and LTE‑Advanced Pro Enhancements

8.10 Case Studies: LTE Deployment, VoLTE, and Performance

Case Study 1: VoLTE (Voice over LTE)

VoLTE uses IMS (IP Multimedia Subsystem) to provide voice services over LTE. It requires QCI 1 for voice bearer (GBR, 50 ms delay). The UE performs IMS registration, and the network sets up dedicated bearers for voice. eSRVCC (Enhanced Single Radio Voice Call Continuity) allows handover to 2G/3G when LTE coverage is lost.

Case Study 2: LTE Capacity Planning

A city with 500,000 subscribers, each generating 0.02 Erlangs, requires a blocking probability of 1%. Using Erlang B, total traffic = 10,000 Erlangs. With 20 MHz bandwidth and 3 sectors per site, the number of eNodeBs is planned based on average throughput per sector (e.g., 50 Mbps).

Case Study 3: LTE‑Advanced in Practice

An operator uses carrier aggregation (20 + 15 MHz) and 4×4 MIMO to achieve peak downlink rates of 450 Mbps. The network uses CoMP for cell‑edge users to improve SINR by 3 dB.

Quiz: 25 Questions

Test your understanding of LTE architecture and operations.

Q1. What are the two main components of the LTE network architecture?

Show Answer

E‑UTRAN (Radio Access Network, comprising eNodeBs) and EPC (Evolved Packet Core, comprising MME, SGW, PGW, etc.).

Q2. What is the role of the MME (Mobility Management Entity) in the EPC?

Show Answer

The MME handles mobility management, authentication, paging, and bearer management (control‑plane functions).

Q3. What is the difference between the S1‑U and S1‑MME interfaces?

Show Answer

S1‑U is the user‑plane interface between eNodeB and SGW; S1‑MME is the control‑plane interface between eNodeB and MME.

Q4. List the layers of the LTE user‑plane protocol stack (from top to bottom).

Show Answer

PDCP → RLC → MAC → PHY (from higher to lower).

Q5. What are the three RLC modes, and what is the main difference between them?

Show Answer

Transparent (TM): no segmentation/ARQ; Unacknowledged (UM): segmentation but no ARQ; Acknowledged (AM): segmentation and ARQ for reliable delivery.

Q6. Which multiple‑access scheme is used in LTE downlink and uplink?

Show Answer

DL: OFDMA (Orthogonal Frequency Division Multiple Access). UL: SC‑FDMA (Single‑Carrier Frequency Division Multiple Access).

Q7. What is HARQ and why is it used in LTE?

Show Answer

Hybrid ARQ combines error correction (FEC) with ARQ retransmissions to improve reliability and efficiency, especially in error‑prone wireless channels.

Q8. What are the two RRC states of a UE in LTE?

Show Answer

RRC_IDLE and RRC_CONNECTED.

Q9. What is the purpose of the Tracking Area (TA) in LTE?

Show Answer

A TA is a group of cells used to manage location of idle UEs. When a UE moves across TA boundaries, it performs a Tracking Area Update (TAU).

Q10. Describe the difference between X2‑based and S1‑based handover.

Show Answer

X2‑based handover uses the direct X2 interface between eNodeBs, avoiding core network involvement. S1‑based handover goes through the MME and involves more core network signalling; used when X2 is not available.

Q11. What is the role of the PDCP layer in LTE?

Show Answer

PDCP provides header compression (ROHC), ciphering (encryption), integrity protection, and in‑order delivery of packets.

Q12. How does LTE perform authentication and key agreement (AKA)?

Show Answer

UE and HSS share a secret key K. HSS generates authentication vectors (RAND, AUTN, XRES, KASME) and sends them to MME. MME sends RAND and AUTN to UE; UE verifies AUTN and sends RES; MME verifies RES against XRES to authenticate.

Q13. What encryption algorithm is commonly used in LTE (besides SNOW 3G)?

Show Answer

AES (Advanced Encryption Standard) with 128‑bit keys.

Q14. What is a QoS Class Identifier (QCI) and what does it specify?

Show Answer

QCI is a scalar value (1‑9) that defines a specific QoS profile, including priority, packet delay budget, and acceptable packet loss rate.

Q15. What is the difference between GBR and Non‑GBR bearers?

Show Answer

GBR (Guaranteed Bit Rate) bearers provide a guaranteed minimum bit rate, used for real‑time services like VoIP. Non‑GBR bearers are best‑effort and share resources.

Q16. What is the purpose of carrier aggregation in LTE‑Advanced?

Show Answer

Carrier aggregation combines multiple component carriers to increase the total bandwidth (up to 100 MHz) and boost data rates.

Q17. What is the typical latency target for LTE user‑plane?

Show Answer

Less than 10 ms (for the radio access network). The end‑to‑end latency is typically around 30‑50 ms.

Q18. Which network element assigns the IP address to the UE in LTE?

Show Answer

The PGW (PDN Gateway) allocates the IP address to the UE as part of the bearer establishment.

Q19. What is the function of the SGW (Serving Gateway)?

Show Answer

The SGW routes and forwards user data packets, acts as a mobility anchor for intra‑LTE handovers, and buffers downlink data for idle UEs.

Q20. What is the role of the HSS in LTE?

Show Answer

HSS (Home Subscriber Server) is the central database containing subscriber profiles, authentication credentials, and service subscription information.

Q21. What is the difference between RRC connection setup and NAS signalling?

Show Answer

RRC is the radio interface protocol that establishes the connection between UE and eNodeB. NAS signalling is between UE and MME (e.g., attach, TAU, authentication) and is carried over RRC in the air interface.

Q22. What is the purpose of the RACH (Random Access Channel) procedure?

Show Answer

RACH is used for initial access, uplink synchronisation, and scheduling requests when the UE has no dedicated resources.

Q23. How does LTE support inter‑RAT handover to GSM/UMTS?

Show Answer

LTE supports inter‑RAT handover via S1 and the MME, which coordinates with the target network's core (e.g., 3G SGSN). The UE performs measurements and reports, and the network triggers the handover.

Q24. What is the maximum number of component carriers supported in LTE‑Advanced for carrier aggregation?

Show Answer

Up to 5 component carriers (in Release 10), each up to 20 MHz, totaling 100 MHz. Later releases support more.

Q25. What is the function of the PCRF in the EPC?

Show Answer

The PCRF (Policy and Charging Rules Function) defines QoS and charging policies based on subscriber profiles and service requirements.

Exercises (12 Practical Problems)

Apply your LTE knowledge to these problems.

Exercise 1: In an LTE network, a UE is in RRC_IDLE mode. Explain the sequence of messages and procedures that occur when the UE initiates a data session (e.g., web browsing).

Show Solution

1. UE performs Random Access (RACH) to obtain synchronisation and a temporary C‑RNTI. 2. UE sends RRC Connection Request (with establishment cause). 3. eNodeB responds with RRC Connection Setup. 4. UE sends RRC Connection Setup Complete (including NAS Service Request). 5. NAS Service Request triggers MME to authenticate the UE (if needed) and set up bearers. 6. MME sends Initial Context Setup Request to eNodeB (with bearer QoS). 7. eNodeB establishes radio bearers (DRBs). 8. Data can now flow.

Exercise 2: Calculate the peak downlink data rate for LTE using 20 MHz bandwidth, 2×2 MIMO, 64‑QAM, and coding rate 5/6. Assume 100 Resource Blocks, each with 12 subcarriers per slot, 7 symbols per slot, 2 slots per subframe, and subframe duration 1 ms. Include 10% overhead.

Show Solution

Data rate = (RB × subcarriers × symbols × bits/symbol × layers × (1/code rate?) Actually, bits per RE = 6 (64‑QAM) × 5/6 = 5 effective bits. Number of REs per subframe = 100 × 12 × 7 × 2 = 16800. Effective bits per subframe = 16800 × 5 × 2 (MIMO) = 168,000 bits. With 1000 subframes/sec (1 ms each), rate = 168 Mbps. With 10% overhead, about 151 Mbps. Standard says 300 Mbps with 4×4 MIMO.

Exercise 3: What is the purpose of the X2 interface? Describe the information exchanged during an X2 handover.

Show Solution

X2 interface connects eNodeBs for direct signalling, enabling fast handover. During X2 handover, the source eNodeB sends a Handover Request to the target eNodeB including UE context, QoS, and security info. Target eNodeB performs admission control and replies with an ACK. Then the source eNodeB forwards user data (buffered) to target. The UE then performs random access to target, and the target signals completion. A Path Switch Request is sent to MME to update SGW.

Exercise 4: A UE is attached to a network with a default EPS bearer. It initiates a VoIP call. Explain the setup of a dedicated bearer for VoIP (QCI 1). What messages are exchanged?

Show Solution

UE sends a Bearer Resource Allocation Request (NAS) to MME. MME interacts with PGW/PCRF to request a dedicated bearer. MME sends E‑RAB Setup Request to eNodeB (with QCI 1, GBR). eNodeB sets up a radio bearer (DRB) with appropriate QoS. The eNodeB replies with E‑RAB Setup Response. The bearer is established, and VoIP traffic can flow.

Exercise 5: Describe the LTE security procedure when a UE attaches for the first time. Include the derivation of keys.

Show Solution

1. UE sends Attach Request (with IMSI or GUTI) to MME. 2. MME sends Authentication Data Request to HSS. 3. HSS derives authentication vectors (RAND, AUTN, XRES, KASME) and sends to MME. 4. MME sends Authentication Request (RAND, AUTN) to UE. 5. UE verifies AUTN, computes RES, and sends Authentication Response (RES). 6. MME verifies RES against XRES. 7. MME derives KASME from K using KDF. 8. MME sends Security Mode Command (with ciphering/integrity algorithms) to eNodeB, which forwards to UE. 9. UE derives CK and IK and responds with Security Mode Complete. 10. Now encryption and integrity are active for subsequent messages.

Exercise 6: What are the differences between RLC UM and RLC AM? When would you use each?

Show Solution

RLC UM (Unacknowledged Mode) provides segmentation and reassembly but no ARQ. It is used for delay‑sensitive services (e.g., VoIP) where retransmissions would add too much delay. RLC AM (Acknowledged Mode) provides ARQ with retransmissions for reliable delivery; used for data services (e.g., TCP) that require low error rates.

Exercise 7: Explain the concept of carrier aggregation and how it increases throughput. What are the types of carrier aggregation (intra‑band, inter‑band)?

Show Solution

Carrier aggregation combines multiple component carriers (CCs) to increase the total bandwidth. Intra‑band aggregation uses carriers within the same band (contiguous or non‑contiguous). Inter‑band aggregation uses carriers from different bands. The UE can receive/transmit on multiple CCs simultaneously, increasing peak data rates.

Exercise 8: How does LTE support mobility in RRC_IDLE mode? What is the paging procedure?

Show Solution

In RRC_IDLE, UE performs cell reselection based on measurements and informs the network of its location via Tracking Area Updates (TAUs) when crossing TA boundaries. For incoming data/voice, the MME sends a paging message to all eNodeBs within the UE's current TA; each eNodeB broadcasts a Paging message over the air. The UE responds with a Service Request to initiate connection setup.

Exercise 9: What is the role of the PDCP layer in handling handover?

Show Solution

During handover, PDCP ensures in‑order delivery of packets. It can perform retransmission (if RLC AM) and reordering after handover. It also handles ciphering and integrity; the keys are updated (via key derivation) during handover.

Exercise 10: Explain the difference between downlink and uplink scheduling in LTE. How does the eNodeB allocate resources?

Show Solution

Downlink scheduling: eNodeB allocates resource blocks (RBs) in the downlink to UEs based on channel quality (CQI), buffer status, and QoS. Uses dynamic scheduling with grants sent on PDCCH. Uplink scheduling: eNodeB grants uplink resources (UL grant) to UEs based on SR (Scheduling Request) and Buffer Status Reports (BSR). The UE transmits on allocated RBs using SC‑FDMA.

Exercise 11: What is the purpose of the S1‑U interface and how does it differ from S1‑MME?

Show Solution

S1‑U is the user‑plane interface between eNodeB and SGW, carrying GTP‑U tunnels for user data. S1‑MME is the control‑plane interface between eNodeB and MME, carrying S1‑AP messages for mobility and bearer management. They are separate to allow independent scaling and security.

Exercise 12: Describe the steps for a VoLTE call setup, including IMS registration and dedicated bearer establishment.

Show Solution

1. UE attaches to LTE and establishes a default EPS bearer. 2. UE performs IMS registration (SIP REGISTER) via the PGW to IMS core. 3. For a call, UE sends SIP INVITE to IMS. 4. IMS triggers PCRF to request a dedicated bearer for voice (QCI 1). 5. PCRF pushes policy to PGW, which initiates a dedicated bearer (via MME/eNodeB). 6. A radio bearer (DRB) is set up with GBR and low latency. 7. Call proceeds with RTP/RTCP streams.

Homework (12 Assignments)

Independent research and advanced analysis.

HW1. Derive the LTE downlink peak data rate formula based on the number of resource blocks, modulation, coding rate, MIMO layers, and overhead. Show how carrier aggregation increases the rate.

Show Sample Answer

Rate = (N_RB × 12 × 14 × bits_per_subcarrier × coding_rate × layers × (1 - overhead) × 1000) / 10^6 Mbps. With N_RB=100, 64‑QAM (6 bits), coding 5/6, 2 layers, overhead 10%, rate ≈ 100 × 12 × 14 × 6 × 0.833 × 2 × 0.9 × 1000 / 1e6 = 100 × 12 × 14 = 16800 REs; 16800 × 6 × 0.833 = 84000 bits per subframe; × 2 = 168000; × 0.9 = 151200 bits/ms = 151.2 Mbps. With carrier aggregation (5 carriers) → 756 Mbps.

HW2. Research and compare the LTE physical layer frame structure (FDD and TDD). How does the subframe configuration differ, and what are the implications for latency and capacity?

Show Sample Answer

LTE FDD uses separate frequency bands for DL and UL, with 10 ms frames (10 subframes). Each subframe is 1 ms. TDD uses the same frequency but splits time between DL and UL, with configurable subframe allocations (e.g., 1:3, 2:2). TDD allows more flexible spectrum use but introduces latency due to switching and may require guard periods. FDD provides consistent latency.

HW3. Explain the role of the S1‑AP protocol in LTE. What are the main procedures (e.g., Initial Context Setup, Handover Preparation)?

Show Sample Answer

S1‑AP is the control‑plane protocol between eNodeB and MME. It handles procedures: Initial UE Message (attach), Initial Context Setup (bearer establishment), Handover Preparation (S1 handover), UE Context Release, Paging, and Tracking Area Update. It uses non‑transparent signalling.

HW4. Analyse the impact of HARQ retransmissions on LTE uplink performance. How does the round‑trip time of HARQ affect latency?

Show Sample Answer

HARQ in LTE uses synchronous UL HARQ (with 8 processes). The RTT is 8 ms (for FDD). Each retransmission adds 8 ms latency. For low‑latency services, this can be significant. Techniques like TTI shortening (in 5G) address this. Also, HARQ with incremental redundancy improves efficiency but adds complexity.

HW5. Describe the key procedures in the LTE attach procedure, including the interaction between UE, eNodeB, MME, HSS, and SGW/PGW. What messages are exchanged?

Show Sample Answer

1. UE sends Attach Request (with IMSI/GUTI) + RRC Connection Setup. 2. eNodeB forwards Attach Request (S1‑AP Initial UE Message) to MME. 3. MME performs authentication with HSS (AKA). 4. MME sends Update Location Request to HSS; HSS replies with subscription data. 5. MME sends Create Session Request to SGW/PGW to establish default bearer. 6. PGW sends Create Session Response with IP address. 7. MME sends Initial Context Setup Request to eNodeB (with bearer QoS). 8. eNodeB sets up radio bearers and sends RRC Connection Reconfiguration to UE. 9. UE responds. 10. eNodeB sends Initial Context Setup Response to MME. 11. MME sends Attach Accept to UE (NAS). 12. Attach Complete.

HW6. Write a critical evaluation of the EPS bearer concept. How does it differ from the PDP context in 3G? What are the advantages?

Show Sample Answer

EPS bearer is a logical connection with QoS attributes, similar to PDP context but more streamlined. In 3G, PDP context is packet‑switched; in LTE, the EPS bearer is the core QoS entity. Advantages: simpler (flat architecture), better QoS granularity with QCI, and integration with policy control (PCRF).

HW7. Research the concept of "SON" (Self‑Organizing Networks) in LTE and LTE‑Advanced. How does it reduce the cost and complexity of network deployment?

Show Sample Answer

SON automates network planning, configuration, optimization, and healing. It includes: self‑configuration (e.g., Automatic Neighbour Relation – ANR), self‑optimization (e.g., mobility robustness optimization, load balancing), and self‑healing (automatic detection and compensation for failures). This reduces manual intervention and operational costs.

HW8. Explain the concept of Coordinated Multi‑Point (CoMP) in LTE‑Advanced. How does it improve SINR at the cell edge? What are the challenges?

Show Sample Answer

CoMP coordinates transmission/reception among multiple cells (eNodeBs). Joint transmission (JT) sends data from multiple cells to a UE, increasing the received signal. Joint reception (JR) combines signals from multiple cells for UL. This improves SINR, especially for edge users. Challenges: high backhaul requirements, tight synchronisation, and increased signalling overhead.

HW9. Analyse the security architecture of LTE, including the key hierarchy (K, KASME, CK, IK, KeNB). How are keys refreshed during handover?

Show Sample Answer

Key hierarchy: K (permanent) → KASME (master key derived after AKA) → KeNB (used for RRC/UP ciphering). During handover, the target eNodeB derives a new KeNB* from the current KeNB using a key derivation function (KDF) with the target cell ID and frequency. This provides forward secrecy.

HW10. Compare the LTE system architecture with the 5G service‑based architecture in terms of flexibility, scalability, and deployment options.

Show Sample Answer

LTE (EPC) uses fixed functional entities and point‑to‑point interfaces, which are less flexible. 5GC uses service‑based architecture (SBA) with network functions (e.g., AMF, SMF) communicating via HTTP/2, enabling modularity, cloud‑native deployment, network slicing, and easier evolution. 5GC is more scalable and supports edge computing, whereas LTE is more monolithic.

HW11. Write a report on the evolution of MIMO in LTE: from 2×2 in Release 8 to 8×8 in Release 10 and beyond, including MU‑MIMO and FD‑MIMO.

Show Sample Answer

LTE introduced MIMO with 2×2 (Release 8). Release 10 added 8×8 DL MIMO and 4×4 UL MIMO. Release 12 added 3D‑MIMO (FD‑MIMO) with active antenna systems. MU‑MIMO (Release 11) allows multiple UEs on the same time‑frequency resources. These improvements increased spectral efficiency and peak rates.

HW12. Design a network planning strategy for LTE deployment in a dense urban area with high user density. Consider spectrum, cell radius, sectorization, and capacity requirements. Include a link budget.

Show Sample Answer

Use 1800 MHz or 2.6 GHz. Cell radius ~500 m (urban). Use 3‑sector sites. Link budget: Tx power 40 dBm, antenna gain 15 dBi, receiver sensitivity -100 dBm, path loss with Okumura‑Hata: ~125 dB at 500m. Fade margin 10 dB. Capacity: assume 50 Mbps per sector, 3 sectors = 150 Mbps per site. With 500 users per km², each generating 0.05 Erlangs, total traffic ≈ 25 Erlangs/km². Number of sites determined by required capacity.

Summary

This extended tutorial provided a comprehensive examination of LTE architecture and operations, covering:

Mastery of LTE principles is essential for understanding modern mobile networks and the transition to 5G. The next tutorial will explore 5G Networks and Emerging Technologies.

© COMP347 – Unit 7: Wireless and Mobile Networks (Extended Tutorial 8)