Improved Generalized Regression Neural Network for Target Localization

Jondhale, Satish R.; Wakchaure, Manoj Ashok; Agarkar, Balasaheb S.; Tambe, Sagar B.

doi:10.1007/s11277-022-09627-9

Cited by 16 publications

(6 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…, deployed in a square area and fixed at known positions, are tested, 6 while the target node (x t , y t ) to be estimated is placed in the area, and it is connected to the computer as the processing center. The RSSI values (RSSI 1 to RSSI 4 ) received by the target node can be converted to the distance values using the log-distance path-loss equation 30,31 as shown in (1) and (2), where RSSI d and RSSI d 0 (in dBm) refer to the measured RSSIs at distance d and the reference distance from the transmitter (d 0 ¼ 1 m).…”

Section: Wcl Methodsmentioning

confidence: 99%

“…The position of an unknown target can be estimated by using the actual reference positions and the distance values as converted from the RSSIs collected from the reference nodes. In our system, four reference nodes,

((),, x_{1}, y_{1})

((),, x_{4}, y_{4})

, deployed in a square area and fixed at known positions, are tested, 6 while the target node (

x_{t}

y_{t}

) to be estimated is placed in the area, and it is connected to the computer as the processing center. The RSSI values (

{RSSI}_{1}

{RSSI}_{4}

) received by the target node can be converted to the distance values using the log‐distance path‐loss equation 30,31 as shown in () and (), where

{RSSI}_{d}

and

{RSSI}_{d_{0}}

(in dBm) refer to the measured RSSIs at distance

d

and the reference distance from the transmitter (

d_{0} = 1 m

).…”

Section: Wcl Methodsmentioning

confidence: 99%

“…The WCL method was proposed by Jondhale et al 6 This method estimates the target position using () to (), where (

x_{est}

y_{est}

) refers the estimated target position,

N

is the number of reference nodes to be used (i.e.,

N = 4

for Figure 2),

w_{i}

is the weight factor calculated from

d_{i}

, and

g

determines the weight of the contribution of the reference node. Here, in general,

g

can be set to 1, as tested by experiments in Blumenthal et al, 10 Wattananavin et al, 14 and Reichenbach and Timmermann 32 .…”

Section: Wcl Methodsmentioning

confidence: 99%

“…Indoor target localization in wireless sensor networks (WSNs) has been one of the most interesting topics in the research community from the past to the present 1–3 . Since position information can be useful in many applications, including human tracking in buildings, patient and elderly tracking in hospitals, mobile robot tracking, worker tracking in construction sites, and location service for shopping malls, 4,5 the design and development of more efficient localization methods are continuously required to fulfill different WSN and Internet of Things (IoT) applications 6–9 . Based on the literature survey, many localization methods have been introduced.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Development of an indoor corridor localization system based on a WCL method with corridor area selection and distance compensation

Wattananavin,

Booranawong,

Saito

2024

Int J Communication

View full text Add to dashboard Cite

SummaryIn this paper, a received signal strength indicator (RSSI)‐based weighted centroid localization (WCL) method for indoor corridor scenarios is developed and tested. The contribution of the proposed system is that, to scope the area of an unknown target position, the possible corridor area where the target node should be located is automatically selected, and only reference nodes deployed in such an area are applied for position estimation. Additionally, to improve the estimation precision, distance values converted from measured RSSIs and used for the WCL are also compensated to alleviate the RSSI variation problem caused by physical environments. Therefore, our methods can save a number of reference nodes to be used, while localization accuracy can also be achieved. Experiments in the corridor environment using a 2.4 GHz IEEE 802.15.4 wireless sensor network with four reference nodes deployed in the test field dimension of 22 m × 9.3 m have been performed. Experimental results demonstrate that the proposed method, which requires only two reference nodes, outperforms the original WCL method with four reference nodes by 53.053%, as indicated by average localization errors. The results also reveal that the proposed method provides error distances lower than the WCL method for all test target positions, while all estimated positions fall within the corridor areas by 100%.

show abstract

Section: Wcl Methodsmentioning

confidence: 99%

((),, x_{1}, y_{1})

((),, x_{4}, y_{4})

, deployed in a square area and fixed at known positions, are tested, 6 while the target node (

x_{t}

y_{t}

) to be estimated is placed in the area, and it is connected to the computer as the processing center. The RSSI values (

{RSSI}_{1}

{RSSI}_{4}

) received by the target node can be converted to the distance values using the log‐distance path‐loss equation 30,31 as shown in () and (), where

{RSSI}_{d}

and

{RSSI}_{d_{0}}

(in dBm) refer to the measured RSSIs at distance

d

and the reference distance from the transmitter (

d_{0} = 1 m

).…”

Section: Wcl Methodsmentioning

confidence: 99%

“…The WCL method was proposed by Jondhale et al 6 This method estimates the target position using () to (), where (

x_{est}

y_{est}

) refers the estimated target position,

N

is the number of reference nodes to be used (i.e.,

N = 4

for Figure 2),

w_{i}

is the weight factor calculated from

d_{i}

, and

g

determines the weight of the contribution of the reference node. Here, in general,

g

can be set to 1, as tested by experiments in Blumenthal et al, 10 Wattananavin et al, 14 and Reichenbach and Timmermann 32 .…”

Section: Wcl Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Development of an indoor corridor localization system based on a WCL method with corridor area selection and distance compensation

Wattananavin,

Booranawong,

Saito

2024

Int J Communication

View full text Add to dashboard Cite

show abstract

“…During tracking, the positioning image of each frame is used to fine-tune the network to adapt it to the actual situation. Environment and Task [2]. Some scholars proposed the FCNT algorithm.…”

Section: Introductionmentioning

confidence: 99%

Algorithm Research of Neural Network for Feature Target Tracking

2022

View full text Add to dashboard Cite

In recent years, with the explosive growth of research in the field of computer vision, TT has an important application in visual recognition tasks. It can assist target detection and improve the speed of recognition, which has certain theoretical value and research significance. However, in practical application scenarios, target tracking (TT) still faces the problems of inaccurate tracking, poor robustness, and low overall system speed caused by scene changes. Since the TT algorithm based on convolutional neural network (CNN) was proposed, it has attracted the attention of a large number of researchers with the advantages of both speed and accuracy. The multi-layer convolution features extracted from the input image through the CNN have a good appearance representation ability for the target under a variety of complex interference factors. Therefore, this paper uses the CNN to build a feature TT algorithm model, and experiments prove that the accuracy of the model can be improved by increasing the order of magnitude and deepening the ResNet backbone network design. After comparing the tracking effects of the CNN algorithm and other algorithms, it shows that the CNN algorithm has the highest tracking accuracy and success rate.

show abstract

Mobility target tracking with meta‐heuristic aided target movement prediction scheme in WSN using adaptive distributed extended Kalman filtering

Ramadevi,

Subramanyam,

Bindu

2024

Int J Communication

View full text Add to dashboard Cite

SummaryIn wireless sensor networks (WSNs), target tracking has been prominently raised in recent days. Because of the frequent utilization of WSN, the attention on target tracking is greatly increased. The estimated optimal value derived from the earlier moment is rarely taken into consideration in traditional target‐tracking algorithms. One of the most crucial uses of WSNs is mobile target tracking, and it is especially used for spying. Precision surveillance is heavily dependent on localization or distance estimation, and extensive study has been done in this area. This research aims to develop a new target‐tracking network‐assisted target movement prediction model in WSN with reduced energy consumption. The major phases involved in the proposed target movement prediction in WSN are (a) mobility target tracking and (b) target movement prediction. Initially, the mobility target tracking is done with the help of adaptive distributed extended Kalman filtering (ADEKF). The mobility target tracking performance is improved by optimally tuning the parameters from ADEKF with the support of the improved squid game optimizer (ISGO). Then, the target movement prediction phase is executed with the help of input parameters like “Angle of Arrival (AoA) and Received Signal Strength (RSS),” and the optimal progress of the mobile node is predicted. The implementation outcome of the proposed tracking network‐assisted target movement prediction model is validated concerning various performance metrics. Overall performance analysis shows that the developed model offers 2.5% in terms of RMSE measures. The developed model shows better analysis while validating with existing approaches.

show abstract

Improved Generalized Regression Neural Network for Target Localization

Cited by 16 publications

References 20 publications

Development of an indoor corridor localization system based on a WCL method with corridor area selection and distance compensation

Development of an indoor corridor localization system based on a WCL method with corridor area selection and distance compensation

Algorithm Research of Neural Network for Feature Target Tracking

Mobility target tracking with meta‐heuristic aided target movement prediction scheme in WSN using adaptive distributed extended Kalman filtering

Contact Info

Product

Resources

About