Introduction to the MAC layer

The LTE RLC and MAC layer is interposed between a PHY, logical processing to achieve physical channels, including channel switching, prioritization and scheduling management. The MAC layer provides the following features:

1, mapping between logical channels and transport channels.

2, from a plurality of logical channels or MAC SDU multiplexed into a transport block (the TB), sent to the physical layer through the transport channel.

3, a plurality of logical channels or MAC SDU demultiplexing the physical layer SDU from the TB is transmitted through a transport channel.

4, the scheduling information reporting.

5, error correction through HARQ.

6, the priority operation performed between the UE through dynamic scheduling.

7, the same operation for priority among logical channels of one UE.

8, the logical channel prioritization.

9, the transmission format selection.

Parsing the MAC layer concept

Channel map

For downlink: multiplexing one or more logical channels from the down data (MAC SDUs) to the transport block and transmits to the physical layer through the transport channel.
For the uplink: the transmission block transmitted from the transmission channel up demultiplexed into MAC SDU, and through the corresponding logical channel, the RLC layer handed.

RNTI

When the MAC indicates the use of radio resources through the PDCCH is a physical channel, MAC logical channel according to the type of RNTI is mapped to the corresponding PDCCH, so that by matching the user can obtain different RNTI corresponding to the logical channel data,

C-RNTI, Temporary C-RNTI and SPS C-RNTI for the DTCH and DCCH;
P-RNTI for the PCCH;
RA-RNTI to receive a random access at the corresponding DL-SCH;
Temporary C-RNTI for receiving a random access procedure in a CCCH;
SI-RNTI used for BCCH.

HARQ

There are two levels of retransmission mechanisms in LTE: HARQ MAC layer, and the layer ARQ RLC (AM Mode). Play a major role in the HARQ MAC layer, the RLC is an ARQ as a supplementary means exist.

HARQ (Hybrid Automatic Repeat Request), hybrid automatic repeat request. Depending on the content of the retransmission, in 3GPP standards and recommendations, mainly three kinds of hybrid automatic retransmission request mechanisms, including HARQ-I, HARQ-II and HARQ-III and the like.
(1) HARQ-I type: FEC Forward Error Correction + retransmit
HARQ-I is the conventional HARQ scheme, it is only the introduction of the FEC coding on the basis of the ARQ.
(2) HARQ-II Type: FEC Forward Error Correction coding in combination + + retransmitted
HARQ-II also called full incremental redundancy scheme. In this embodiment wherein the retransmitted data is not simply replicated data has been transmitted, but the additional redundant information. Receiving end decodes every combination, the combinations of all the previously received bit codeword is formed of a lower rate, so that a greater coding gain can be obtained, to achieve incremental redundancy purposes. Each time the amount of redundant retransmission are different, and the retransmission data can not be decoded alone, can only be merged with the data previously transmitted in order to be decoded.
(3) HARQ-III type: FEC Forward Error Correction + + retransmission complementary deleted

For each data packet transmitted using a complementary manner deleted, each data packet may be decoded individually, may be encoded into one packet having a larger redundant information merge decoded.

TTI bundling

The TTI : Transmission Time interval The transmission time interval length which illustrates a MAC transport block on the time, in LTE TTI = 1ms.

Bundling the TTI : transmission on a plurality of multiple consecutive subframes with a TB (Transport Block), without waiting for ACK / NACK in the art.

TTI bundling purpose is to improve the uplink VoIP coverage of a cell edge UE. According to some known simulation results, the use of TTI bundling can bring up 4 dB of gain. No use 3GPP downlink TTI BUNDING set forth.

Why propose TTI bundling?

Different UE, different maximum transmit power, for the edge of the UE, uplink coverage is small, the data transmission error rate can not meet the acceptable range. In uplink data transmission in the normal (non-TTI-bunding), TB will be converted into a plurality of redundancy versions RV, the UE first transmits a frame at a sub-RV, RV whether the subsequent transmission, one ACK / NACK front depends. UE edges, a higher probability of data retransmission, resulting in a larger delay voiP business experience decline.

In the TTI bundling mechanism, corresponding to the same TB may transmit different RV (4 uplink subframes) in successive sub-frames without the need to wait for a response ACK / NACK. When all the transmission and reception corresponding to TB are dealt with, it will send a joint ACK / NACK. That is, after transmission of the same TB frame receiving a plurality of times (different RV) in consecutive sub, and make the soft combining process, using one ACK / NACK unify response.

TTI bundling open mechanism

eNodeB when to enable a UE TTI bundling it? One implementation is the eNodeB a given period of time, by receiving a corresponding UE Power whether the UE to calculate the available power headroom below a certain threshold (for example: a transmission power close to the maximum transmit power of the UE, but still very low SINR value), to determine whether to enable TTI bundling function. Or it is found in the cell boundary UE: When (e.g. very low SINR) and the UE using similar voice services (QCI = 1), so TTI bundling is enabled on the UE.

Precautions

Only FDD and TDD 0/1/6, supports TTI bundling. For the other four TDD UL, a system since the uplink sub-frame number is less than 4, is not supported TTI bundling.
For TDD, it is not at the same time enabling TTI bundling and the SPS.
If the UE is configured with one or more SCell upstream (not downstream SCell), the UE can configure the TTI bundling. FDD and TDD are applicable, i.e., does not support the uplink carrier aggregation TTI bundling)
MSG3 transmitting the random access procedure is not used in the TTI bundling.
If the UE using TTI bundling, the UE can be allocated in a TTI within 3 RB at most, only the uplink data is transmitted using QPSK;
All of the radio bearers used by the UE if the TTI bundling, then the UE (DRB includes not only for voice transmission) can only use TTI bundling.
All TTI TTI transmissions within the bundle as a whole, a uniform feedback HARQ ACK / NACK. I.e., only the bundle TTI corresponding to the last TTI, will receive a HARQ ACK / NACK feedback, regardless of whether the transmission data TTI.
TTI bundle retransmission is still a TTI bundle.

BSR

Buffer Status Report (BSR), used to provide a buffer status report to the serving eNB UE has a total of presence information in the buffer upstream of how much data needs to be sent. There are several BSR: regular BSR, periodic BSR and a padding BSR.

Conventional BSR: no uplink resource allocation, when a logical channel belongs to a logical channel group has data to transmit, the need to trigger regular BSR.

Padding BSR: the uplink resources have been allocated, and the number of bits is greater than or equal to the padding BSR control information but plus its sub-header, this time triggered BSR filled.

Periodic BSR: When periodicBSR the Timer- the BSR timer expires trigger called "periodic BSR".

SR

Scheduling request (SR), for requesting uplink shared channel resources used for transmitting the uplink data. Only regular BSR will be SR. When triggered SR, UE to send a buffer status report (BSR), BSR sent on the shared channel, need resources to send, then how to obtain uplink resources for sending the BSR it? This first sending SR on the PUCCH or send via PRACH. But if the SR sent on PUCCH always fail, then it is necessary to obtain uplink resources by way of random access.

Random access

A random access procedure is to establish an uplink radio link between the UE and the network, and only after the random access is completed before normal data transfer between the eNB and the UE. Depending on the service trigger, random access can be divided into a contention-based random access (Contention based random access procedure), and based on the non-contention random access (Non-Contention based random access procedure ).

Non-contention random access of the UE according to an instruction of the eNB, using a random access code assigned Preamble initiated on the specified PRACH channel resources.

UE contention based random access is no uplink resources when there are data to be transmitted, using the random code Preamble initiating random access.

MAC data format

A MAC PDU of a MAC header (MAC header) + 0 or more MAC SDU + 0 or more MAC CE (Control Element) + padding may be present composition. MAC header of a MAC subheader or more components. Each subheader corresponds to a MAC PDU, or a MAC CE, or padding.

è¿éåå¾çæè¿ °

MAC header of a variable length, which comprises the following parameters:

LCID: it is used to indicate a logical channel, a control message type field or filled;
L: SDU or indicating the message length of the control, except the last sub-header and a control message corresponding to the fixed length header, each head has a sub-field L, which length is indicated by the field F;
F: If the length is greater than the SDU or control messages 128Byte, then set F = 1, otherwise set to 0, a value of F, we can know the size of the corresponding L value, i.e. know the content (SDU or the MAC Control unit length of the message);
E: Indicates whether there is a plurality of MAC header fields, when E = 1, means that there is another set of next R / R / E / LCID field, if it is 0, then the next step is the payload;
R: Reserved bit, set to "0"