MAC Layer Overview in LTE 4G

Introduction

The LTE MAC is the lowest layer of Layer 2, sitting directly between RLC above and the physical layer below. It is the sublayer that turns "several kinds of data that want to send" into "one transport block on the shared radio, delivered reliably and on time," and it is specified in TS 36.321.

It is active whenever the UE is in RRC_CONNECTED and moving data, and its random-access machinery even runs from RRC_IDLE to bootstrap a connection. Every user-plane packet and every signalling message crosses the MAC on its way to and from the air: RLC hands it PDUs, the MAC multiplexes them into a transport block with any control elements, and the PHY carries it. Nothing reaches the radio without passing through here.

It matters because the air interface is a single shared, error-prone medium, and the MAC is where the hard real-time arbitration happens — which flow gets the grant, how corrupted blocks are retransmitted in milliseconds, how the uplink stays time-aligned, and how the UE sleeps to save battery. Understanding the MAC is understanding how LTE actually shares the radio.

Why the MAC Layer Is Needed

The radio is one shared, error-prone medium and the physical layer sends exactly one transport block per direction each millisecond. That creates four problems that some layer must solve, and the MAC is the lowest layer that still reasons about which data goes out. It must decide how competing logical flows share that one block; it must correct air errors fast enough that TCP and voice do not stall; it must keep every UE's single-carrier uplink time-aligned at the eNB; and it must let an idle-ish UE stop listening to save battery. Everything below the MAC only knows how to code and modulate bits; everything above it thinks in streams and does not know or care about the shared radio. The MAC is the seam that reconciles the two.

What the LTE MAC layer actually does

Everything above the MAC thinks in terms of logical channels — streams of a particular kind of information (signalling on DCCH, user data on DTCH, paging on PCCH). Everything below the MAC thinks in terms of transport channels — chunks of bits the physical layer knows how to code and modulate (DL-SCH, UL-SCH, PCH, BCH, RACH). The MAC's core job is to bridge those two worlds and to arbitrate the one thing every UE in a cell must share: the radio.

The MAC offers data-transfer and radio-resource-allocation services to RLC above it, and uses the transport-channel services of the physical layer below. Its functions are best enumerated as a checklist, because in an LTE exam or design review you are expected to be able to name every one of them and say what it buys you.

Logical channels and transport channels

The two channel families are the vocabulary of the MAC, so it is worth pinning down exactly what each one is. A logical channel is defined by what it carries — the type of information — and lives at the RLC/MAC boundary. A transport channel is defined by how the information is carried over the air — the format, coding and delivery characteristics the physical layer applies — and lives at the MAC/PHY boundary. The MAC is precisely the layer where a logical channel is mapped onto a transport channel.

Logical channels split into control channels (carrying signalling from the control plane) and traffic channels (carrying user-plane data). The full LTE set, including the MBMS channels used for broadcast/multicast services, is below.

Transport channels describe the delivery service: whether the data is broadcast to everyone or addressed to one UE, whether it supports HARQ, dynamic link adaptation, beamforming and so on. LTE has a compact set.

The MAC's mapping is largely fixed: BCCH maps to BCH (for the MIB) or DL-SCH (for SIBs); PCCH maps to PCH; CCCH, DCCH and DTCH map to DL-SCH downlink and UL-SCH uplink; MCCH and MTCH map to MCH. The RACH carries no logical channel — it is a pure physical-access resource. See Channel Mapping for the complete matrix and how it differs slightly between control and user plane.

The MAC entity, transport blocks and RNTIs

In basic LTE a UE has a single MAC entity; from Release 12 onward, when dual connectivity is configured, it has one MAC entity per cell group — one for the Master Cell Group (MCG) and one for the Secondary Cell Group (SCG). Each entity independently runs its own HARQ, DRX, BSR and random-access machinery for the cells it serves. With carrier aggregation the entity serves the PCell plus one or more SCells, each of which can be activated or deactivated by MAC signalling.

The rhythm of the LTE MAC is the TTI: every 1 ms subframe, the MAC delivers one transport block per TTI in each direction (a second block is added only for downlink spatial multiplexing with two codewords). The scheduler decides the transport block size, and the MAC's job is to fill exactly that many bits — no more, no less. This fixed 1 ms TTI is a defining LTE trait: the whole retransmission and scheduling timeline is clocked to it, unlike NR where the slot length shrinks with numerology.

On the shared channel there is no dedicated pipe per UE, so the network needs a way to say "this grant is for you." That is the job of the RNTI — a Radio Network Temporary Identifier that scrambles the CRC of the PDCCH carrying the grant. The UE's persistent identity in a cell is its C-RNTI (Cell RNTI, a 16-bit value in the range 0x003D–0xFFF3); a PDCCH whose CRC unmasks with the UE's C-RNTI is a grant addressed to that UE, and the UE finds it by blind-decoding its search spaces on PDCCH.

Where the MAC sits and what flows through it

The MAC lives directly below RLC and directly above the physical layer. RLC hands it PDUs (which become MAC SDUs); the MAC multiplexes them with any control elements into a transport block and passes it to PHY over a transport channel. Read the figure top to bottom: logical channels enter from RLC, the MAC's internal functions (LCP/mux, HARQ, random access, DRX/TA) act on them, and one transport block per TTI leaves on a transport channel to PHY.

Multiplexing, the MAC PDU and HARQ

When the scheduler grants the UE resources, the MAC must fill exactly the granted number of bits. It runs LCP to decide the order in which logical channels contribute their SDUs, then assembles a MAC PDU: a MAC header followed by MAC SDUs, MAC control elements and, if needed, padding. Each element is described by a subheader carrying an LCID in the header. MAC CEs carry in-band control and are placed before the MAC SDUs in the PDU. See MAC PDU & CEs for the byte-level layout.

Two things about LTE multiplexing matter when you compare it to NR. First, in LTE the RLC layer does concatenation: RLC can glue several higher-layer SDUs into one RLC PDU sized to fit the grant, so the MAC's multiplexing job is mostly about combining different logical channels and control elements, not sizing a single flow. Second, because RLC does that sizing, the LTE MAC learns the grant size and then instructs RLC how many bytes each logical channel may contribute this TTI — a tightly coupled MAC/RLC interaction each subframe.

LCP gives each logical channel a priority and a Prioritized Bit Rate (PBR). In the first round each channel is served up to its PBR in strict priority order, so a high-priority channel cannot completely starve a lower one; any grant left over is then filled by strict priority until the grant is exhausted or the buffers are empty. This is how the UE splits one grant fairly across, say, a signalling DCCH and a data DTCH.

The MAC also owns a small set of in-band MAC control elements, each identified by a reserved LCID so the receiver can tell control from data. They travel inside the same MAC PDU as the user data, with no separate signalling channel — the reason routine control never has to climb up to RRC.

Scheduling reports, random access, timing advance and DRX

The eNB scheduler cannot allocate uplink well unless it knows two things: how much data is waiting, and how much transmit power is left. The MAC reports both as CEs — the BSR and the PHR. If the UE has data but no grant on which to send a BSR, it fires a Scheduling Request (SR) on PUCCH to ask for one; if it has no SR resource configured, it falls back to random access. This BSR/SR/PHR trio is the entire uplink feedback loop that keeps the scheduler fed.

Random access is the procedure a UE uses when it has no uplink grant and, often, no valid timing alignment — at initial access, after handover, when re-establishing after radio-link failure, or when uplink data arrives while unsynchronised. Contention-based access is the familiar four-step exchange: the UE sends a preamble on the RACH (Msg1); the eNB replies with a Random Access Response addressed by RA-RNTI, granting an uplink resource and an initial timing advance (Msg2); the UE sends Msg3 on UL-SCH carrying its identity (a C-RNTI MAC CE or CCCH message); and the eNB resolves contention with Msg4, echoing the UE's identity in a Contention Resolution Identity MAC CE. A contention-free variant (a dedicated preamble assigned by the network) skips the contention step for handover and re-synchronisation. See Random Access.

Because LTE is single-carrier on the uplink (SC-FDMA), every UE's signal must arrive at the eNB roughly time-aligned regardless of its distance. The MAC maintains this with timing advance (TA): it applies each Timing Advance Command to shift its transmit timing earlier and starts the timeAlignmentTimer; while that timer runs the uplink is considered synchronised, and if it expires the UE flushes its HARQ buffers, releases PUCCH/SRS resources, and must go back through random access to re-synchronise before it may transmit again.

Once connected, blindly monitoring PDCCH every subframe would drain the battery. DRX lets the MAC switch the receiver off on a schedule. During the On Duration the UE monitors PDCCH; if nothing is scheduled it may sleep until the next cycle, and any grant restarts the drx-InactivityTimer to keep it awake while traffic flows. LTE defines a long and an optional short DRX cycle, and a DRX Command MAC CE can send the UE to sleep early. See DRX for the timer details.

Carrier aggregation and SCell activation (LTE-A)

LTE-Advanced lets a UE receive on several component carriers at once — carrier aggregation. One MAC entity serves the anchor PCell plus up to several SCells, each an additional carrier. The PCell is always active and carries the essential control (the PUCCH, security and mobility anchor); SCells are added by RRC but can be switched on and off dynamically without RRC signalling, because MAC does the switching.

The lever is the Activation/Deactivation MAC CE. A bitmap in that CE marks which configured SCells are active. Activation is fast — MAC-speed rather than RRC-speed — so the network can add bandwidth the moment a burst arrives and drop it (via deactivation or an sCellDeactivationTimer expiry) to save UE power when the burst ends. A deactivated SCell is still configured but the UE does not monitor its PDCCH, does not transmit SRS, and reports no CQI for it. Each active SCell has its own HARQ entity, so aggregation multiplies transport-block throughput while the single MAC entity keeps one scheduling, BSR and DRX view across all of them.

How LTE MAC differs from NR MAC

If you already know NR, the LTE MAC will feel familiar but subtly different, and interviewers love the contrasts. The functions are broadly the same — mapping, multiplexing, HARQ, LCP, scheduling reports, random access, DRX — but several structural choices diverge.

The single most quoted difference is HARQ: LTE's synchronous uplink HARQ versus NR's fully asynchronous scheme in both directions. The second is that LTE relies on RLC concatenation, which means the LTE MAC multiplexing story is genuinely different from NR's — in NR, RLC no longer concatenates, so the layers were re-split to shave latency. And NR's BWP machinery simply has no LTE equivalent: an LTE UE always works across the full configured carrier. Keep those three in your pocket and the rest of the comparison follows.

Summary

The LTE MAC (TS 36.321) is Layer 2's lowest sublayer, bridging the logical channels RLC hands down (BCCH, CCCH, DCCH, DTCH, PCCH, MCCH, MTCH) to the transport channels PHY carries (DL-SCH, UL-SCH, PCH, BCH, MCH, RACH). Its job is to arbitrate one shared, error-prone medium: it multiplexes several flows into one transport block per 1 ms TTI via LCP, corrects air errors with HARQ, keeps the uplink fed with BSR/SR/PHR, opens connections with random access, holds the single-carrier uplink time-aligned via timeAlignmentTimer, and saves battery with DRX.

Grants are addressed by RNTI (the persistent C-RNTI plus special ones for RA, SI, paging, SPS), and in-band MAC CEs carry routine control without troubling RRC. LTE-A adds fast SCell activation for carrier aggregation via a MAC CE. Against NR, the enduring contrasts are LTE's synchronous UL HARQ, its reliance on RLC concatenation, and its lack of bandwidth parts. Get the channel vocabulary and those three contrasts straight and the rest of the MAC follows.