Efficient way of Non-GBR, High Latency GTP-U Packet Transmission in 4G and 5G Networks

In Telecom Network as per 3GPP proposed architecture and protocol stack, GTP-U protocol is used for the packet transmission between Nodes (Serving Gateway (SGW), PDN Gateway (PGW), SGW-U, PGWU, eNode-B, UPF, gNode-B). As per GTP-U protocol Tunnel Endpoint Identifier (TEID) allocated at every node for a PDU session of a UE. TEID is unique identifier based on which the telecom nodes identify UE's PDU Session and process them. Thus, for every data packet of UE GTP-U header is added over the payload when it travels between the GTP-U nodes. There can be packets belong to 1 to 10 thousand UEs that can be transmitted between these GTPU node at a single point of time. Irrespective of high latency or low latency packets as soon as the GTP-U node receives the payload it adds the GTP-U header and sends out to the peer node. The problem over here is ineffective bandwidth utilization due to this mechanism of transmission for Non-Guaranteed Bit Rate (Non-GBR) and high latency application related packets. IP header and UDP header which will be same for every GTP-U packet that will be transmitted between the GTP-U nodes. For low latency application packets cannot be buffered and hence there is no other best suited way of GTP-U packet transmission. High latency application packets also transmitted without buffering which results in underutilization of bandwidth. Our novel algorithm here solves the problem of underutilization of bandwidth by buffering and combining the high latency packets of different UEs' PDU sessions that needs to be transmitted from a GTP-U node towards the target GTP-U node.