2010
DOI: 10.1109/tvlsi.2009.2017542
|View full text |Cite
|
Sign up to set email alerts
|

Detailed Characterization of Transceiver Parameters Through Loop-Back-Based BiST

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
5

Citation Types

0
14
0

Year Published

2011
2011
2022
2022

Publication Types

Select...
4
3
1

Relationship

1
7

Authors

Journals

citations
Cited by 43 publications
(14 citation statements)
references
References 18 publications
0
14
0
Order By: Relevance
“…However, these technologies are prone to higher process variations and defect rates, which makes RF BIST both more necessary and more challenging. Several techniques in the literature aim to reduce the dependence on external RF instrumentation for low cost testing [2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18]. In [2][3][4], simple test signals such as multi-tone sinusoidal signals, are used to generate output data, where the performance of RF transceiver is predicted using the output data as well as machine learning methods.…”
Section: Introductionmentioning
confidence: 99%
See 1 more Smart Citation
“…However, these technologies are prone to higher process variations and defect rates, which makes RF BIST both more necessary and more challenging. Several techniques in the literature aim to reduce the dependence on external RF instrumentation for low cost testing [2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18]. In [2][3][4], simple test signals such as multi-tone sinusoidal signals, are used to generate output data, where the performance of RF transceiver is predicted using the output data as well as machine learning methods.…”
Section: Introductionmentioning
confidence: 99%
“…In [5][6][7][8][9][10], the loopback configuration is used to characterize both the transmitter and the receiver. In [5,6], the authors derive the analytical model for the entire loop-back path and use numerical techniques in order to solve transmitter and receiver parameters simultaneously. In [7], authors use similar mathematical models with specialized signals in order to achieve an analytical solution.…”
Section: Introductionmentioning
confidence: 99%
“…Especially for RF circuits this objective is hard to satisfy since these circuits are very sensitive and tapping into their signal paths may seriously unbalance the performance trade-offs achieved by design. For example, loop-back test for transceivers [6], [7], [8], [9], where the test signals are generated in the baseband and the transmitter's output is connected to the receiver's input This paper is an invited summary paper on the Ph.D. dissertation work of Louay Abdallah [I] that was earried out at TIMA Laboratory (CNRSGrenoble INP -UJF), Grenoble, Franee, from Oetober 2008 to Oetober 2012 under the supervision of Haralampos-G. Stratigopoulos and Salvador Mir: Parts of this summary paper have been previously published in [2] , [3], [4] , [5]. This summary paper was invited in the framework of the finals of the 2013 IEEE Computer Soeiety Test Teehnology Teehnieal Couneil (TTTC) doetoral thesis award competition that took plaee at the 20 to analyze the test response also in the baseband, requires the insertion of a switch and an attenuator to perform the connection and, for some types of transceivers, even an extra mixer is inserted in the RF signal path.…”
Section: Introductionmentioning
confidence: 99%
“…Recently, in [12], Erdogan and Ozev describe a method for characterizing the RF transceiver parameters from analysis of loopback data obtained for a specified number of "random" OFDM frames. In addition to the fundamental nonlinearity parameters diagnosed in [3,11], they are also [13,14] it is shown how a few multitones can be used to determine RF system level as well as embedded module performance metrics using nonlinear solvers as discussed in [7][8][9][10]12]. A key benefit of this test/diagnosis methodology is that it does not require the use of supervised learners as used in [1][2][3].…”
Section: Introductionmentioning
confidence: 99%
“…A key benefit of this test/diagnosis methodology is that it does not require the use of supervised learners as used in [1][2][3]. In the current work, optimized multitones are used for test stimulus as opposed to the random OFDM frames used in [12]. A key requirement for tuning is that the test response must be maximally sensitive to process as well as tuning knob perturbations across the allowed ranges of tuning knob values and across multiple process corners.…”
Section: Introductionmentioning
confidence: 99%