Search

CN-122028091-A - Bridging method and system for inference and assurance of reliable communication for large-scale connection

CN122028091ACN 122028091 ACN122028091 ACN 122028091ACN-122028091-A

Abstract

The invention discloses a bridging method and a bridging system for inference and assurance of large-scale connection reliable communication, comprising the steps of jointly inferring an active terminal set and a channel thereof from pilot frequency observation, and generating posterior samples and confidence information; and converting the posterior sample and the confidence information into a provable signal power confidence lower bound and a residual item upper bound to form an equivalent signal-to-interference-plus-noise ratio confidence lower bound, and combining an interference power consistency upper bound generated by a space regularity condition to output a reliability certificate. The invention upgrades the system from 'higher performance' to 'performance-proof, controllable and auditable'.

Inventors

  • Tang Chenyue
  • YANG HAO

Assignees

  • 浙江大学

Dates

Publication Date
20260512
Application Date
20260414

Claims (10)

  1. 1. A method of bridging inference and assurance of reliable communications, comprising: Jointly deducing an active terminal set and a channel thereof from pilot frequency observation, and generating a posterior sample and confidence information; converting the posterior sample and the confidence information into a provable signal power confidence lower bound and a residual item upper bound to form an equivalent signal-to-interference-and-noise ratio confidence lower bound; and outputting a reliability certificate by combining the interference power consistency upper bound generated by the space regularity condition.
  2. 2. The method of claim 1, wherein jointly inferring the set of active terminals and their channels from pilot observations and generating posterior samples and confidence information comprises: Score-generated posterior inference model and pilot observations With pilot set To input and output active information Channel point estimation for each terminal And calculate the error confidence radius Wherein For terminal indexing, K is the total number of candidate terminals, Is a confidence level parameter.
  3. 3. The method according to claim 1, characterized in that the signal power confidence lower bound is translated, in particular: let the received combined vector From estimating the channel The real effective signal amplitude is constructed as , wherein, Taking the modulus of the vector and superscript Represents the conjugate transpose of the object, The uplink channel vector is the kth terminal; At a confidence event The following uses the triangle inequality and the cauchy-schwarz inequality: ; Wherein the method comprises the steps of As a function of the vector norm, As a radius of the confidence in the error, Is a confidence level parameter; thereby yielding a lower bound for signal power in the sense of confidence: 。
  4. 4. the method of claim 1, wherein the residual term upper bound is: ; wherein alpha 1 ,α 2 is more than or equal to 0 as a coefficient, Is the confidence radius of the error, wherein For terminal indexing, K is the total number of candidate terminals, Is a confidence level parameter.
  5. 5. The method according to claim 1, characterized by letting For the true external interference power of link k in the current slot, For thermal noise power, a lower signal power bound is given under confidence events And by Describing the upper bound of the residual term, the confidence lower bound of the equivalent signal to interference plus noise ratio The method comprises the following steps: 。
  6. 6. The method of claim 5, wherein the interference power generated in conjunction with the spatial regularity condition is consistent with an upper bound Defining auditable lower bounds under spatial regularity : ; Defining a reliability certificate : ; Where θ is the reliability threshold, Is a confidence level parameter.
  7. 7. The method of claim 6, wherein Φ is defined as Stable point process of all potential transmitting nodes on the table To take the following measures Is the center and radius If the finite constants sigma is more than or equal to 0, rho is more than or equal to 0 and v is more than or equal to 0, the following events occur with probability 1: ; Then phi is said to satisfy the strong sphere regularity, where (sigma, p, v) is referred to as the normalized intensity parameter, Representation of all And all of The intersection is taken from the condition set; Setting a path loss function Is a non-negative, bounded, monotonically non-increasing function and satisfies And is also provided with ; At any receiving position Defining a geometric aggregate path loss sum generated by potential transmitting nodes: ; Wherein, the Representing a point process A node location of (a); For truncated radius, representing statistical distance-only receiving position No more than Node contribution of (a); Representation of all And is also provided with Node position of (2) Summing, i.e. split balls The potential transmitting nodes in the network are overlapped item by item; When Φ meets the strong sphere rule, there is a definite upper bound for all R > 0: Let R → infinity obtain the computable upper bound constant of the total space aggregate path loss sum : ; Thereby to arbitrary Has the following components ; Assuming that the transmitting power of each active transmitting node on the same frequency RB satisfies 0≤P x ≤P max , the aggregate interference power of any receiving point satisfies the upper bound of consistency: 。
  8. 8. a bridging system for inference and assurance of reliable communications, comprising: Generating a push layer, namely jointly deducing an active terminal set and a channel thereof from pilot frequency observation, and generating a posterior sample and confidence information; The bridging layer is used for converting the posterior sample and the confidence information into a provable signal power confidence lower bound and a residual item upper bound to form an equivalent signal-to-interference-and-noise ratio confidence lower bound; And a guarantee layer, namely outputting a reliability certificate by combining an interference power consistent upper bound generated by a space regularity condition.
  9. 9. A computer readable storage medium having stored thereon a program, which when executed by a processor is adapted to carry out a reliable communication inference and assurance bridging method according to any one of claims 1 to 7.
  10. 10. A computer program product comprising a computer program which, when executed by a processor, implements a reliable communication inference and assurance bridging method as claimed in any one of claims 1 to 7.

Description

Bridging method and system for inference and assurance of reliable communication for large-scale connection Technical Field The invention relates to the technical field of wireless communication and artificial intelligence intersection, in particular to a bridging method and a bridging system for inference and assurance of reliable communication for large-scale connection. Background In a massive terminal random access scenario (mMTC), a base station needs to complete closed loop decision of 'whether a terminal is active, why channel parameters are, and whether target reliability is met' simultaneously under the conditions of extremely short pilot frequency and strong interference. On the one hand, the existing access side method generally outputs statistical indexes such as detection accuracy, mean square error and the like. Even though the average performance is stronger, the average meaning is always better, the available and auditable reliability guarantee is difficult to be given to a certain time slot and a certain specific link, when the training data is inconsistent with the statistical rule of the real deployment environment, the algorithm can be degraded, and the system lacks quantification and protection means for the degradation degree. On the other hand, the reliability analysis at the network side usually characterizes the node position randomness by means of a Poisson Point Process (PPP), the model is convenient for obtaining results such as average success rate and the like, but the model naturally describes average performance of typical links, and is difficult to directly ensure that all links are established simultaneously, and more importantly, when space control conditions are lacked, the total interference is difficult to be consistent and up-bound in a nearly everywhere sense, so that the reliability lower bound established for all links lacks a provability foundation. Therefore, what is lacking in engineering is a cross-layer bridging mechanism that can not only utilize the advantages of generating type inference under complex channel distribution, but also inject inference errors and uncertainties into a reliability lower bound framework in a computable and provable manner, so that the system not only gives an estimation result, but also can output a reliability certificate which can be used for admittance, scheduling and retransmission. The invention proposes an "inference and guaranteed bridging" scheme for the gap. Disclosure of Invention The invention aims at overcoming the defects of the prior art and providing a bridging method and a bridging system for inference and assurance of reliable communication of large-scale connection. The invention aims at realizing the following technical scheme that the method for bridging the inference and the assurance of reliable communication comprises the following steps: Jointly deducing an active terminal set and a channel thereof from pilot frequency observation, and generating a posterior sample and confidence information; converting the posterior sample and the confidence information into a provable signal power confidence lower bound and a residual item upper bound to form an equivalent signal-to-interference-and-noise ratio confidence lower bound; and outputting a reliability certificate by combining the interference power consistency upper bound generated by the space regularity condition. Further, jointly inferring the set of active terminals and their channels from pilot observations and generating posterior samples and confidence information, comprising: Score-generated posterior inference model and pilot observations With pilot setTo input and output active informationChannel point estimation for each terminal(I.e., a posterior sample) and calculate the error confidence radius(I.e., confidence information). Further, the signal power confidence lower bound is converted specifically as follows: let the received combined vector From estimating the channelThe real effective signal amplitude is constructed as, wherein,Is a complex modulus, superscriptRepresents the conjugate transpose of the object,The uplink channel vector is the kth terminal; At a confidence event The following uses the triangle inequality and the Cauchy-Schwarz inequality: ; Wherein the method comprises the steps of As a vector norm (default euro norm),As a radius of the confidence in the error,Is a confidence level parameter; thereby yielding a lower bound for signal power in the sense of confidence: wherein Is a non-negative truncation operator. Further, the residual term upper bound is: ; wherein alpha 1,α2 is more than or equal to 0 as a coefficient, Is the confidence radius of the error, whereinFor terminal indexing, K is the total number of candidate terminals,Is a confidence level parameter. Further, let theFor the true external interference power of link k in the current slot,For thermal noise power, a lower signal power bound is given under confidence event