Dynamic packet buffering duration

Abstract

Certain aspects of the present disclosure provide techniques for packet buffering. A method that may be performed by a receiving node includes dynamically determining one or more time durations to buffer packets. The one or more time durations can be different than a time duration of a configured timer for buffering the packets. The receiving node may input one or more parameters to a machine learning algorithm and obtain, as output of the machine learning algorithm based on the input one or more parameters, one or more time durations to buffer packets. The receiving node buffers packets for the determined one or more time durations. The receiving node may use machine learning to dynamically determine the one or more time durations to buffer packet. The buffering may be at a radio link control (RLC) reassembling buffer and/or a packet data convergence protocol (PDCP) buffer.

1 . An apparatus for wireless communication, comprising:
 a memory; and   at least one processor coupled with the memory, the memory and at least one processor configured to:
 input one or more parameters to a machine learning algorithm; 
 obtain, as output of the machine learning algorithm based at least in part on the input one or more parameters, one or more time durations to buffer packets, the one or more time durations being different than a time duration of a configured timer for buffering the packets; and 
 buffer the packets for one of the one or more time durations. 
   
     
     
         2 . The apparatus of  claim 1 , wherein at least one of the one or more time durations is shorter than the time duration of the configured timer. 
     
     
         3 . The apparatus of  claim 1 , wherein at least one of the one or more time durations is longer than the time duration of the configured timer. 
     
     
         4 . The apparatus of  claim 1 , wherein:
 the configured timer comprises a radio link control (RLC) reassembly timer or reordering timer; and   the at least one processor is configured to buffer the packets at an RLC layer of the apparatus.   
     
     
         5 . The apparatus of  claim 1 , wherein:
 the configured timer comprises a packet data convergence protocol (PDCP) reordering timer; and   the at least one processor is configured to buffer the packets at a PDCP layer of the apparatus.   
     
     
         6 . The apparatus of  claim 1 , wherein the one or more parameters further includes historical values associated with the one or more parameters. 
     
     
         7 . The apparatus of  claim 1 , wherein the one or more parameters comprise one or more lower layer block error rates (BLERs). 
     
     
         8 . The apparatus of  claim 1 , wherein the one or more parameters comprise one or more numbers of hybrid automatic repeat request (HARQ) retransmissions used to determine a HARQ delay. 
     
     
         9 . The apparatus of  claim 1 , wherein the one or more parameters comprise one or more numbers of radio link control (RLC) retransmissions used to determine an RLC delay. 
     
     
         10 . The apparatus of  claim 1 , wherein the one or more parameters comprise one or more dual connectivity configurations of the apparatus. 
     
     
         11 . The apparatus of  claim 1 , wherein the one or more parameters comprise one or more of: reordering timer expiry history, reassembly timer expiry history, one or more numbers of times one or more missing packets are received, one or more prior time durations at which one or more missing packets are received, a minimum prior time duration at which one or more missing packets are received, a maximum prior time duration at which one or more missing packets are received, one or more logical channel identifiers (LCIDs) associated with the packets, amount of memory left for buffering, one or more radio resource control (RRC) configurations, one or more evolved universal mobile telecommunications system (UMTS) terrestrial radio access (EUTRA) new radio (NR) dual connectivity (ENDC) configurations, one or more split bearer configurations, overall CPU utilization, clock frequency, numerology, downlink (DL) transport block (TB) size, uplink (UL) grant size, a maximum number of hybrid automatic repeat request (HARQ) retransmissions used, a HARQ round trip time (RTT), a time taken to transmit an uplink status protocol data unit (PDU), a histogram of holes receives in one or more time duration bins, a throughput for a radio bearer associated with the one or more packets, a traffic type associated with the one or more packets, a delay between dual connectivity links, a tune away time, a gain state, an average signal to noise ratio (SNR), geo-location information, carrier information, a number of active component carriers, transport block size, average packet data convergence protocol (PDCP) packet size, a time division duplex (TDD) configuration, a frequency division duplex (FDD) configuration, an application profile, a radio bearer mode, a single subscriber identify module (SSIM) configuration, a multiple SIM (MSIM) configuration, a modem operation conditions, application data protocol, a quality-of-service (QoS) profile associated with the application, or a combination thereof. 
     
     
         12 . The apparatus of  claim 11 , wherein one or more of the one or more parameters are per carrier parameters. 
     
     
         13 . The apparatus of  claim 1 , wherein the memory and at least one processor are further configured to:
 enable the machine learning algorithm for use based at least in part on reaching a threshold rate of the machine learning algorithm successfully predicting time durations for receiving missed packets.   
     
     
         14 . The apparatus of  claim 1 , wherein the memory and at least one processor are further configured to:
 enable the machine learning algorithm for use, per radio bearer, based at least in part on an application type attached to the radio bearer.   
     
     
         15 . The apparatus of  claim 14 , wherein the memory and at least one processor are configured to obtain, as the output of the machine learning algorithm based at least in part on the input one or more parameters, the one or more time durations to buffer packets including a time duration, per radio bearer, to buffer packets. 
     
     
         16 . The apparatus of  claim 1 , wherein the memory and at least one processor are configured to:
 input to the machine learning algorithm a first set of parameters for a first link with a first node;   input to the machine learning algorithm a second set of parameters for a second link with a second node;   obtain, as the output from the machine learning algorithm, a first time duration for the first link; and   obtain, as the output from the machine learning algorithm, a second time duration for the second link.   
     
     
         17 . The apparatus of  claim 16 , wherein the memory and at least one processor are further configured to:
 determine a packet, of a sequence of packets, is missing;   input the first set of parameters when the missing packet is associated with the first link; and   input the second set of parameters when the missing packet is associated with the second link.   
     
     
         18 . The apparatus of  claim 1 , wherein the memory and at least one processor are further configured to:
 input second one or more parameters to a second machine learning algorithm to determine a second one or more time durations; and   select the one or more time durations as opposed to the second one or more time durations to buffer packets.   
     
     
         19 . The apparatus of  claim 1 , wherein the memory and at least one processor is configured to obtain, as the output of the machine learning algorithm based at least in part on the input one or more parameters, the one or more time durations to buffer packets including a predicted time duration to wait for a specified portion of missing protocol data units (PDUs). 
     
     
         20 . The apparatus of  claim 19 , wherein the portion of missing PDUs comprises a number of packets to maintain a maximum application throughput. 
     
     
         21 . The apparatus of  claim 1 , wherein the memory and at least one processor are further configured to:
 obtain, as the output of the machine learning algorithm, a probability, for each of the one or more time durations, of receiving one or more missing packets, and   select the one of the one or more time durations based, at least in part, on the probability.   
     
     
         22 . The apparatus of  claim 1 , wherein the apparatus comprises a user equipment (UE) or a base station (BS). 
     
     
         23 . The apparatus of  claim 1 , wherein the memory at least one processor are further configured to:
 detect a missing packet;   initiate the configured timer; and   halt the buffering after the one of the one or more time durations.   
     
     
         24 . The apparatus of  claim 1 , wherein the memory and at least one processor are further configured to:
 detect a missing packet;   determine an updated timer duration based on the one of the one or more time durations; and   initiate the configured timer with the updated timer duration.   
     
     
         25 . The apparatus of  claim 1 , wherein the memory and at least one processor are further configured to after buffering the packets for the one of the one or more time durations:
 flush a first protocol layer buffer containing the buffered packets; and   send the buffered packets to a second protocol layer, wherein the first protocol layer is a lower protocol layer than the second protocol layer.   
     
     
         26 . The apparatus of  claim 1 , wherein the memory and at least one processor are further configured to:
 after buffering the packets for the one of the one or more time durations, send a radio link control (RLC) status packet data unit (PDU), from a first protocol layer to a second protocol layer, wherein the RLC status PDU indicates one more missing PDUs for retransmission, and wherein the second protocol layer is a lower layer than the first protocol layer.   
     
     
         27 . The apparatus of  claim 1 , wherein the memory and at least one processor are further configured to obtain, as output of the machine learning algorithm, additional one or more time durations to buffer packets at one or more different times. 
     
     
         28 . A method for wireless communications by a node, comprising:
 inputting one or more parameters to a machine learning algorithm;   obtaining, as output of the machine learning algorithm based at least in part on the input one or more parameters, one or more time durations to buffer packets, the one or more time durations being different than a time duration of a configured timer for buffering the packets; and   buffering the packets for one of the one or more time durations.   
     
     
         29 . An apparatus for wireless communications, comprising:
 means for inputting one or more parameters to a machine learning algorithm;   means for obtaining, as output of the machine learning algorithm based at least in part on the input one or more parameters, one or more time durations to buffer packets, the one or more time durations being different than a time duration of a configured timer for buffering the packets; and   means for buffering the packets for one of the one or more time durations.   
     
     
         30 . A computer readable medium storing computer executable code thereon for wireless communications by a node, comprising:
 code for inputting one or more parameters to a machine learning algorithm;   code for obtaining, as output of the machine learning algorithm based at least in part on the input one or more parameters, one or more time durations to buffer packets, the one or more time durations being different than a time duration of a configured timer for buffering the packets; and   code for buffering the packets for one of the one or more time durations.   
     
     
         31 .- 36 . (canceled)