Bandwidth Reservation Accuracy in RSVP Protocols
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional RSVP protocols face accuracy loss and overflow issues due to the use of 32-bit integer type for bandwidth reservation, leading to inaccurate reservations in high-traffic scenarios, especially when dealing with Gbps or Tbps magnitudes.
Innovation Solution
The method extends the resource reservation protocol by including additional accuracy information, such as data type and unit information, in the resource reservation request and confirmation messages, allowing for more precise bandwidth reservation by using float type and extended fields in the FLOWSPEC object, preventing accuracy loss and overflow.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If 32-bit integer type is used for bandwidth reservation, then the protocol structure is simple, but accuracy loss occurs due to insufficient bit length
Solution Approach 1:
The patent extends the bandwidth reservation from a single 32-bit integer field to a multi-dimensional structure consisting of an integer part and a fractional part, each with specific bit allocations. This dimensional expansion allows precise representation of bandwidth values up to Tbps magnitude while maintaining protocol efficiency.
Solution Approach 2:
The patent changes the data representation parameters by introducing a fractional bit field and redefining the bandwidth calculation formula. The bandwidth is now calculated as (integer part) + (fractional part / 2^fraction_bits), transforming the reservation mechanism from simple integer storage to a precision-controlled numerical representation.
2Device complexity
If 32-bit integer type is used for bandwidth reservation, then the configuration is simple, but overflow occurs when bandwidth reaches Gbps or Tbps magnitude
Solution Approach 1:
The patent segments the bandwidth value into two separate fields: an integer part field and a fractional part field. This segmentation allows the system to represent large bandwidth values without overflow, as the combined precision of both fields exceeds the capacity of a single 32-bit integer, while keeping each individual field manageable in size.
3Measurement precision
If float type is used for bandwidth representation, then accuracy is improved, but conversion from integer type causes accuracy loss
Solution Approach 1:
The patent creates a precise copy of the bandwidth value by storing both the integer and fractional parts separately in the message. This copying mechanism ensures that the complete precision information is preserved during transmission, avoiding the accuracy loss that would occur during integer-to-float conversion in conventional systems.
Data Source
AI summary
A resource reservation method, a message sending method, a message receiving method, and a node apparatus in the field of network communication technology are provided. The resource reservation method includes the following steps. A resource reservation request message is received. A resource reservation value and additional accuracy information of the resource reservation value are recorded in the message. Bandwidth resources are reserved for a source node according to the additional accuracy information and the resource reservation value. Thus, resources can be reserved accurately and traffic loss caused by accuracy loss of reservation resources in the case of high traffic is avoided.