Time-Synchronized Resource Reservation in Packet Switched Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current resource reservation mechanisms in packet-switched networks, such as those defined by IEEE 802.1Qat and 802.1Qav standards, do not allow for time-correlated scheduling of resource reservations, leading to delays and inefficiencies in time-sensitive applications like professional video production and broadcast environments.
Innovation Solution
Incorporating time information into resource reservation requests and database management by core network switches to synchronize resource reservations, ensuring that reservations and releases are aligned with a shared wall clock, thereby minimizing delays and maintaining strict quality of service (QoS) by using a method that includes adding a timestamp to resource reservation protocol data units (PDUs) and storing this information for accurate scheduling.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If resource reservation mechanisms are implemented without time correlation, then resource reservation can be established, but the end-time of current reservation cannot be indicated and future reservations cannot be scheduled
Solution Approach 1:
The patent applies preliminary action by including time information (start time and duration) in resource reservation requests before the actual reservation period begins. This allows the network switch to pre-calculate and pre-allocate resources for the specified time period, enabling accurate scheduling of both current and future reservations without waiting for real-time requests.
Solution Approach 2:
The patent introduces time information as an intermediary element that mediates between the resource reservation request and the network switch's resource allocation database. This time information acts as a key parameter that enables the switch to correlate reservations temporally, determine end-times, and schedule future reservations by comparing timestamps against current time.
2Adaptability or versatility
If packet switched networks are used to replace point-to-point infrastructure, then network flexibility is improved, but latencies impact overall system performances
Solution Approach 1:
The patent applies preliminary action by pre-calculating latency values and incorporating them into resource reservation requests before traffic actually flows through the network. Network switches use these pre-provided latency values to allocate resources and schedule transmissions in advance, compensating for packet switched network latencies and ensuring deterministic delivery times for time-sensitive applications.
3Productivity
If resource reservations are made without time information, then reservation establishment is simplified, but delays occur in resource allocation and release
Solution Approach 1:
The patent applies preliminary action by including time information (start time and duration) in resource reservation requests before the actual reservation period begins. This allows the network switch to pre-calculate and pre-allocate resources for the specified time period, enabling accurate scheduling of both current and future reservations without waiting for real-time requests.
Solution Approach 2:
The patent replaces the mechanical system of real-time resource allocation with a time-based automated scheduling system. By substituting manual or reactive resource management with algorithmic time-correlated allocation, the system eliminates delays associated with real-time decision-making and enables proactive resource preparation.
Data Source
AI summary
A method is intended for scheduling reservations of resources in a packet-switched communication network comprising end-nodes (H1-H3) and core network switches (S1-S2). This method comprises the steps of:including a time information to each request, issued by an end-node (H1) and forwarded by at least one core network switch (S1), and requesting a resource reservation for receiving a stream defined by specifications at a time which is represented by this time information, andstoring the time information, which is included into each forwarded end-node request in correspondence with the associated stream specifications, into at least one related database managed by a core network switch (S1) which participates to the request forwarding.


