NR Protocol Stack: PDCP

Functions of PDCP:

1. Header Compression
2. Ciphering and Integrity protection
3. Routing and duplication of split bearers
4. In sequence delivery

Header Compression

For small packets like voice packets or acknowledgment packets, the header itself as long as the data packets which is a huge overhead.

The PDCP performs header compression to reduce the header length to a couple of bytes before wireless transmission and it decompresses the header in the receiving site before it is transmitted over IP protocols over the wired network.

The header compression is based on robust header compression, which is a standardized compression framework that is also used for other mobile communication technologies.

The header compression/decompression serves the purpose of reducing the effective no of bits transmitted over the radio interface.

Ciphering and Integrity protection

Prevent Eavesdropping — There should not be someone who can decode the information transmitted, other than the receiver.

To counter this, the PDCP employs ciphering.

Ciphering is the process of taking on encrypted information called plain text and encrypting it into ciphertext using the encryption key.

The key text in NR is generated using 5 inputs and an encryption algorithm called NEA (NR Encryption Algorithm).

The most important input is the key for RRC encryption which is the security key used for the algorithm. It also uses inputs like the bearer ID, PDCP count, etc.

Data Origin — To verify the data originated from the right source, it performs integrity protection.

In this process, code is added for the identification of the source. For this purpose, the NR uses a similar system using an algorithm called NIA (NR Integrity Algorithm).

It uses a similar set of inputs and creates a MAC authentication code, which is appended to the message.

In the input to the key generation, we have Key RRC integrity which is the important input for the integrity protection process.

Routing and Duplication

In this system, a UE is simultaneously connected to the 2 serving base stations, either both NR or 1 NR and 1 LTE.

The cell that is connected to the core network is called the master cell and the cell that provides additional radio resources is called the secondary cell.

In order to send the data over both the base stations to the UE, split bearers are used where the data is split into 2 data streams. One is transmitted over the master cell and the other is transmitted over the secondary cell. The 2 streams can be combined in the primary cell or the secondary cell.

It is the PDCP that is responsible for routing the data flows from the different bearers to master and secondary cell groups (another cell tower).

In addition to using dual connectivity to achieve more data rate, duplication can be used to send the exact same data through different radio bearers of 2 different cells to improve reliability. In this way, if one of the parts gets corrupt, the information can be retrieved from the other data bearer.

In such cases, the PDCP is responsible for duplicating the information between the different radio bearers. On the receiving side, the PDCP is responsible for discarding the duplicates and picking the un-erroneous packets if there is an error on one of the radio bearers.

This ability to select the multiple copies from the receiving side is called selection diversity and it improves the reliability.

In-sequence vs Out-of-sequence delivery

In some applications, like file transferring even if the packets don’t arrive in the exact same sequence, it doesn’t matter. The TCP which functions above the IP can take care of reordering the packets. The packets that are arrived in the PDCP layer are immediately forwarded to the higher layers.

For some applications like video streaming, the packets have to arrive in the exact sequence as transmitted. For such cases, PDCP is configured for in-sequence delivery. In such cases, on the transmitter side, PDCP adds a sequence no to insure in-sequence delivery.

The basic idea in in-sequence delivery is that the received PDU is stored in the buffer until all the previous packets are received. So only when the packets have arrived, then we forward both to higher layers. This way we ensure that the higher layers receive packets in the exact order as they are transmitted.