Method for the conjoint communication and sensing of the environment of a network node
View Patent ↗The present invention relates to a conjoint communication and sensing method for detecting the movement of an obstacle in the environment of a node in a wireless telecommunication network and, where applicable, localising this obstacle. The node divides its environment into angular sectors and measures, at a plurality of observation instants, a quantity characteristic of the signal-to-noise-plus-sector-interference ratio in each of these sectors ( 410 - 420 ) to form a sensing matrix (Γ τ,n (t)). A blind separation of sources is implemented to extract from this matrix a matrix of sector contributions ( Γ τ , n ( · ) ( t ) ) of the interfering signals ( 430 ). The movement of any obstacle is detected from the history of the sector contributions ( 440 ), which makes it possible, where applicable, to predict a blockage situation ( 450 ) and to modify the allocation of the radio resources accordingly ( 460 ).
1 . Conjoint communication and sensing method for sensing a movement of an obstacle in an environment of a node, referred to as a reference node, of a wireless communication network, said reference node having previously established a radio link with a sending node of the network, said method comprising:
determining a plurality of non-overlapping angular sectors in said environment and vertices of which are the reference node, as well as a plurality of observation instants;
measuring, in each of the observation instants, a measurement of a quantity characteristic of a signal-to-noise-plus-sector-interference ratio on said radio link, for each of the angular sectors, and constructing a sensing matrix
(
Γ
τ
,
n
(
t
)
,
Γ
τ
,
n
j
0
(
t
)
)
,
the elements of which are the measurements of said characteristic quantity;
implementing a blind separation of sources using said sensing matrix and extracting, therefrom, a matrix of sector contributions
(
Γ
τ
,
n
(
·
)
(
t
)
,
Γ
τ
,
n
j
0
(
·
)
(
t
)
)
of interfering signals in the plurality of angular sectors; and
detecting the movement of the obstacle in the environment of the reference node from at least one history of said sector contributions.
2 . Conjoint communication and sensing method according to claim 1 , wherein a situation of blockage of said radio link is predicted from characteristics of a movement of the obstacle in the environment of the reference node.
3 . Conjoint communication and sensing method according to claim 1 , wherein, if such a situation of blockage of said radio link is predicted, a modification of radio resources in the network is implemented.
4 . Conjoint communication and sensing method according to claim 1 , wherein the blind separation of sources is implemented by means of a singular value decomposition of the sensing matrix or a principal components analysis.
5 . Conjoint communication and sensing method according to claim 1 , wherein the reference node determines nodes of its neighbourhood and makes a simultaneous mapping and localisation on said environment using the matrix of sector contributions that it has extracted as well as matrices of sector contributions that it has received, in the form of messages, from the nodes of its neighbourhood.
6 . Conjoint communication and sensing method according to claim 5 , wherein a mapping dictionary is constructed and updated, each entry in the dictionary corresponding to intersection of a plurality of angular sectors the vertices of which are the reference node or nodes of its neighbourhood, each entry containing a signature consisting of the sector contributions of the angular sectors involved in the intersection associated with said entry.
7 . Conjoint communication and sensing method according to claim 6 , wherein, in event of modification of at least one signature in the mapping dictionary between two observation instants, a change is detected and a reallocation of radio resources of the network is implemented.
8 . Conjoint communication and sensing method according to claim 6 , wherein a Kullback-Leibler divergence is calculated between distribution of the signatures at a current observation instant and that of a previous observation instant, over the whole of the mapping dictionary or a part thereof, and in that a reallocation of radio resources of the network is implemented when the Kullback-Leibler divergence exceeds a predetermined threshold.
9 . Conjoint communication and sensing method according to claim 5 , wherein the neighbourhood of the reference node is obtained by making a preselection of nodes of the network by means of a random sampling or a K-nearest neighbours method, then calculates, for each node thus preselected, a score from its matrix of sector contributions and that of the reference node, the neighbourhood of this node being determined by the preselected nodes the score of which exceeds a predetermined threshold or by the M nodes having the highest scores, where M is a non-zero integer.
10 . Conjoint communication and sensing method according to claim 9 , wherein the score of a preselected node is calculated by means of a cosine similarity metric, a Kullback-Leibler divergence or an attention mechanism or by a neural network implementing a classification operation.