Standardized hot-pluggable transceiving unit, hosting unit and method for applying delays based on port positions
Networking device and method using a neural network for applying delays based on port positions. The networking device comprises a plurality of ports and applies a delay to IP packets received via a given port among the plurality of ports. The delay is based on a position of the given port among the plurality of ports. The neural network has been trained to infer the delay to be applied to the IP packet received via the given port. In one implementation, the position of the given port is one of the inputs of the neural network.
1. A networking device comprising:
memory for storing a predictive model comprising weights of a neural network;
a plurality of ports; and
a processing unit for:
receiving an Internet Protocol (IP) packet via a given port among the plurality of ports;
executing a neural network inference engine using the predictive model for calculating an inferred delay based on inputs, the inputs comprising a position of the given port among the plurality of ports and one or more additional input parameter;
applying the inferred delay to the IP packet received via the given port among the plurality of ports;
receiving a new IP packet via the given port among the plurality of ports;
determining that a variation of at least one among the one or more additional parameter is not greater than a given threshold; and
applying the inferred delay to the new IP packet received via the given port among the plurality of ports.
2. The networking device of claim 1 , wherein another IP packet is received via another port among the plurality of ports, another inferred delay is calculated by the neural network inference engine based on the inputs comprising a position of the other port among the plurality of ports and the one or more additional input parameter, and the other inferred delay is applied to the other IP packet received via the other port among the plurality of ports.
3. The networking device of claim 1 , wherein a new IP packet is received via the given port among the plurality of ports, a determination is made by the processing unit that a variation of at least one among the one or more additional parameter is greater than the given threshold, a new inferred delay is calculated by the neural network inference engine based on the inputs comprising the position of the given port among the plurality of ports and the one or more additional input parameter, and the new inferred delay is applied to the new IP packet received via the given port among the plurality of ports.
4. The networking device of claim 1 , wherein the one or more additional input parameter comprises at least one of the following: a time during the day, a maximum throughput of the given port, a current throughput of the given port, an aggregate maximum throughput of all the ports among the plurality of ports for which the neural network inference engine is used for calculating delays, and an aggregate current throughput of all the ports among the plurality of ports for which the neural network inference engine is used for calculating delays.
5. The networking device of claim 1 , wherein at least some among the plurality of ports are adapted for receiving transceiving units, the transceiving units comprising a housing adapted for insertion into the at least some among the plurality of ports.
6. The networking device of claim 5 , wherein at least some among the transceiving units are standardized hot-pluggable transceiving units comprising a housing having standardized dimensions.
7. The networking device of claim 1 , wherein the received IP packet transports a video payload.
8. A method for applying delays based on port positions, the method comprising:
storing in a memory of a networking device a predictive model comprising weights of a neural network;
receiving by a processing unit of the networking device an Internet Protocol (IP) packet via a given port among a plurality of ports of the networking device;
executing by the processing unit of the networking device a neural network inference engine using the predictive model for calculating an inferred delay based on inputs, the inputs comprising a position of the given port among the plurality of ports and one or more additional input parameter;
applying by the processing unit of the networking device the inferred delay to the IP packet received via the given port among the plurality of ports;
receiving by the processing unit of the networking device a new IP packet via the given port among the plurality of ports;
determining by the processing unit of the networking device that a variation of at least one among the one or more additional parameter is not greater than a given threshold; and
applying by the processing unit of the networking device the inferred delay to the new IP packet received via the given port among the plurality of ports.
9. The method of claim 8 , wherein another IP packet is received by the processing unit of the networking device via another port among the plurality of ports of the networking device, another inferred delay is calculated by the neural network inference engine based on the inputs comprising a position of the other port among the plurality of ports of the networking device and the one or more additional input parameter, and the other inferred delay is applied by the processing unit of the networking device to the other IP packet received via the other port among the plurality of ports of the networking device.
10. The method of claim 8 , wherein a new IP packet is received by the processing unit of the networking device via the given port among the plurality of ports of the networking device, a determination is made by the processing unit of the networking device that a variation of at least one among the one or more additional parameter is greater than the given threshold, a new inferred delay is calculated by the neural network inference engine based on the inputs comprising the position of the given port among the plurality of ports and the one or more additional input parameter, and the new inferred delay is applied by the processing unit of the networking device to the new IP packet received via the given port among the plurality of ports of the networking device.
11. The method of claim 8 , wherein the one or more additional input parameter comprises at least one of the following: a time during the day, a maximum throughput of the given port, a current throughput of the given port, an aggregate maximum throughput of all the ports among the plurality of ports for which the neural network inference engine is used for calculating delays, and an aggregate current throughput of all the ports among the plurality of ports for which the neural network inference engine is used for calculating delays.
12. The method of claim 8 , wherein at least some among the plurality of ports of the networking device are adapted for receiving transceiving units, the transceiving units comprising a housing adapted for insertion into the at least some among the plurality of ports of the networking device.
13. The method of claim 12 , wherein at least some among the transceiving units are standardized hot-pluggable transceiving units comprising a housing having standardized dimensions.
14. The method of claim 8 , wherein the received IP packet transports a video payload.