Article Contents
Article Contents

# New optimal error-correcting codes for crosstalk avoidance in on-chip data buses

• * Corresponding author: Xiande Zhang

The first author is supported by the Chinese Scholarship Council at USTC, China. The second author is supported by NSFC grant 11771419, and by the Fundamental Research Funds for the Central Universities

• Codes that simultaneously provide for low power dissipation, cross-talk avoidance, and error correction in the ultra deep submicron/nanometer VLSI fabrication, were recently introduced by Chee et al. in 2015. Such codes were revealed to be closely related to balanced sampling plans avoiding adjacent units, which are widely used in the statistical design of experiments. In this paper, we construct a new family of optimal codes with such properties, by determining the maximum size of packing sampling plans avoiding certain units.

Mathematics Subject Classification: Primary: 05B07, 05B40; Secondary: 94B25.

 Citation:

• Figure 1.  Leave graphs of LPSA$(6n,3;2)$ having $3n$ edges

Figure 2.  Points of the hole are denoted by $\bullet$ and others by $\circ$

Figure 3.  Points of the hole are denoted by $\bullet$ and others by $\circ$

Table 1.  Types of worst crosstalk couplings

 Type-Ⅰ Type-Ⅱ Type-Ⅲ Type-Ⅳ $0\longleftrightarrow1$ $001\longleftrightarrow010$ $010\longleftrightarrow101$ $001\longleftrightarrow110$ $010\longleftrightarrow100$ $011\longleftrightarrow100$ $011\longleftrightarrow101$ $101\longleftrightarrow110$ Single wire undergoes transition. Adjacent wires maintain previous states Center wire in opposite transition to an adjacent wire. The other wire in same transition as center wire Center wire in opposite transition to an adjacent wire. The other wire maintains previous state All three adjacent wires undergo opposite transitions

Table 2.  Upper bounds and leave graphs of CPSAs and LPSAs excluding edges within distance two

 Sampling plans Upper bounds Leave graphs CPSA$(6n,3;2)$ $6n(n-1)$ A perfect matching CPSA$(6n+1,3;2)$ $n(6n-3)-2$ A cycle of length four LPSA$(6n,3;2)$ $6n(n-1)+1$ see Fig. 1 LPSA$(6n+1,3;2)$ $n(6n-3)$ A single edge

Table 3.  Existence results of small orders

 $n$ $24$ $25$ $30$ $31$ $36$ $37$ $48$ $49$ $B^{\circ}(n,3;2)$ $-$ $82$ $-$ $133$ $-$ $196$ $336$ $358$ $B(n,3;2)$ $73$ $84$ $121$ $135$ $181$ $198$ $337$ $360$

Table 4.  Parameters for proof of Lemma 4.6

 n=3g+t-1 g t s 72s=3(24(s-1)+18)+18 24(s-1)+18 19 $s\geq 1, \ s\neq 1$ 72s+24=3(24(s-1)+18)+42 24(s-1)+18 43 $s\geq 1, \ s\neq 1,2,3$ 72s+48=3(24s+12)+12 24s+12 13 $s\geq 1$

Table 5.  Parameters for proof of Lemma 4.8

 $n\quad \ \ \ \ \ \ =3g+t$ $g$ $t$ $s$ $72s+1 \ =3(24(s-1)+18)+19$ $24(s-1)+18$ $19$ $s\geq 1,\ s\neq 1$ $72s+7 \ =3(24s)+7$ $24s$ $7$ $s\geq 1$ $72s+13=3(24(s-1)+18)+31$ $24(s-1)+18$ $31$ $s\geq 1,\ s\neq 1,2,3$ $72s+19=3(24s+6)+1$ $24s+6$ $1$ $s\geq 1$ $72s+25=3(24(s-1)+18)+43$ $24(s-1)+18$ $43$ $s\geq 1, \ s\neq 1,2,3$ $72s+31=3(24s+6)+13$ $24s+6$ $13$ $s\geq 1, \ s\neq 1$ $72s+37=3(24(s-1)+12)+1$ $24(s-1)+12$ $1$ $s\geq 1, \ s\neq 1$ $72s+43=3(24s+6)+25$ $24s+6$ $25$ $s\geq 1, \ s\neq 1$ $72s+49=3(24s+12)+13$ $24s+12$ $13$ $s\geq 1$ $72s+55=3(24s+18)+1$ $24s+18$ $1$ $s\geq 1$ $72s+61=3(24s+18)+7$ $24s+18$ $7$ $s\geq 1$ $72s+67=3(24s+6)+49$ $24s+6$ $49$ $s\geq 1, \ s\neq 1,2,3$
•  [1] D. Bertozzi, L. Benini and G. D. Micheli, Low power error resilient encoding for on-chip data buses, Proceedings of the Conference on Design, Automation and Test in Europe, IEEE Computer Society, (2002), 102–109. doi: 10.1109/DATE.2002.998256. [2] J. A. Bondy and U. S. R. Murty, Graph Theory with Applications, American Elsevier Publishing Co., Inc., New York, 1976. [3] D. Bryant, Y. X. Chang, C. A. Rodger and R. Wei, Two-dimensional balanced sampling plans excluding contiguous units, Communications in Statistics-Theory and Methods, 31 (2002), 1441-1455.  doi: 10.1081/STA-120006078. [4] Y. M. Chee, C. J. Colbourn and A. C. H. Ling, Optimal memoryless encoding for low power off-chip data buses, 2006 IEEE/ACM International Conference on Computer Aided Design, IEEE, (2006), 369–374. doi: 10.1145/1233501.1233575. [5] Y. M. Chee, C. J. Colbourn, A. C. H. Ling, H. Zhang and X. D. Zhang, Optimal low-power coding for error correction and crosstalk avoidance in on-chip data buses, Des. Codes Cryptogr., 77 (2015), 479-491.  doi: 10.1007/s10623-015-0084-4. [6] C. J. Colbourn and A. C. H. Ling, A class of partial triple systems with applications in survey sampling, Communications in Statistics-Theory and Methods, 27 (1998), 1009-1018.  doi: 10.1080/03610929808832141. [7] C. J. Colbourn and A. C. H. Ling, Balanced sampling plans with block size four excluding contiguous units, Australasian Journal of Combinatorics, 20 (1999), 37-46. doi: ajc.maths.uq.edu.au. [8] C. J. Duan, A. Tirumala and S. P. Khatri, Analysis and avoidance of cross-talk in on-chip buses, HOT 9 Interconnects Symposium on High Performance Interconnects, IEEE, (2001), 133–138. doi: 10.1109/HIS.2001.946705. [9] M. Favalli and C. Metra, Bus crosstalk fault-detection capabilities of error-detecting codes for on-line testing, IEEE Transactions on Very Large Scale Integration (VLSI) Systems, 7 (1999), 392-396.  doi: 10.1109/92.784100. [10] A. S. Hedayat, C. R. Rao and J. Stufken, Sampling plans excluding contiguous units, Sampling, Handbook of Statist., North-Holland, Amsterdam, 6 (1988), 575-583.  doi: 10.1016/0378-3758(88)90070-5. [11] A. Hedayat, C. Rao and J. Stufken, 24 Designs in survey sampling avoiding contiguous units, Handbook of Statistics, 6 (1988), 575-583.  doi: 10.1016/S0169-7161(88)06026-2. [12] I. Iqbal, M. H. Tahir, M. Akhtar, S. S. A. Ghazali, J. Shabbir and N. S. Bukhari, Generalized polygonal designs with block size 3 and $\lambda = 1$, Journal of Statistical Planning and Inference, 139 (2009), 3200-3219.  doi: 10.1016/j.jspi.2009.02.018. [13] Z. Khan, T. Arslan and A. T. Erdogan, A dual low power and crosstalk immune encoding scheme for system-on-chip buses, International Workshop on Power and Timing Modeling, Optimization and Simulation, (2004), 585–592. doi: 10.1007/978-3-540-30205-6_60. [14] H. R. Kong, J. G. Lei and Y. Zhang, On constructions for two dimensional balanced sampling plan excluding contiguous units with block size four, Discrete Mathematics, 308 (2008), 3729-3743.  doi: 10.1016/j.disc.2007.07.065. [15] F. J. MacWilliams and N. J. A. Sloane, The Theory of Error-Correcting Codes. I, North-Holland Mathematical Library, Vol. 16. North-Holland Publishing Co., Amsterdam-New York-Oxford, 1977. [16] Y. Miao and L. Zhu, Existence of incomplete group divisible designs, J. Comb. Math. Comb. Comput., 6 (1989), 33-49. [17] K. N. Patel and I. L. Markov, Error-correction and crosstalk avoidance in DSM busses, Proceedings of the 2003 International Workshop on System-level Interconnect Prediction, ACM, (2003), 9–14. doi: 10.1145/639929.639933. [18] S. Ramprasad, N. R. Shanbhag and I. N. Hajj, A coding framework for low-power address and data busses, IEEE Transactions on Very Large Scale Integration (VLSI) Systems, 7 (1999), 212-221.  doi: 10.1109/92.766748. [19] D. Rossi, V. E. S. van Dijk, R. P. Kleihorst, A. H. Nieuwland and C. Metra, Coding scheme for low energy consumption fault-tolerant bus, Proceedings of the Eighth IEEE International On-Line Testing Workshop (IOLTW), IEEE, (2002), 8–12. doi: 10.1109/OLT.2002.1030176. [20] J. E. Simpson, Langford sequences: Perfect and hooked, Discrete Mathematics, 44 (1983), 97-104.  doi: 10.1016/0012-365X(83)90008-0. [21] M. R. Stan and W. P. Burleson, Bus-invert coding for low-power I/O, IEEE Transactions on Very Large Scale Integration (VLSI) Systems, 3 (1995), 49-58.  doi: 10.1109/92.365453. [22] J. Stufken, Combinatorial and statistical aspects of sampling plans to avoid the selection of adjacent units, J. Combin. Info. Syst. Sci., 18 (1993), 149-160. [23] J. Stufken and J. H. Wright, Polygonal designs with blocks of size $k \leq 10$, Metrika, 54 (2001), 179-184.  doi: 10.1007/s001840100137. [24] C.-L. Su, C.-Y. Tsui and A. M. Despain, Saving power in the control path of embedded processors, IEEE Design Test of Computers, 11 (1994), 24-31.  doi: 10.1109/54.329448. [25] M. H. Tahir, I. Iqbal, M. Akhtar and J. Shabbir, Cyclic polygonal designs with block size 3 and $\lambda = 1$ for joint distance $\alpha = 6$ to $16$, Journal of Statistical Theory and Practice, 4 (2010), 203-220.  doi: 10.1080/15598608.2010.10411981. [26] B. Victor and K. Keutzer, Bus encoding to prevent crosstalk delay, Proceedings of the 2001 IEEE/ACM International Conference on Computer-Aided Design, (2001), 57–63. doi: 10.1109/ICCAD.2001.968598. [27] X. M. Wang, T. Feng, J. Zhang and Y. Xu, Two-dimensional balanced sampling plans avoiding adjacent units, Discrete Mathematics, 338 (2015), 1624-1642.  doi: 10.1016/j.disc.2015.04.012. [28] J. H. Wright, Two-dimensional balanced sampling plans excluding adjacent units, Journal of Statistical Planning and Inference, 138 (2008), 145-153.  doi: 10.1016/j.jspi.2007.05.016. [29] J. H. Wright and J. Stufken, New balanced sampling plans excluding adjacent units, Journal of Statistical Planning and Inference, 138 (2008), 3326-3335.  doi: 10.1016/j.jspi.2006.10.020. [30] J. Zhang and Y. X. Chang, The spectrum of cyclic BSEC with block size three, Discrete Mathematics, 305 (2005), 312-322.  doi: 10.1016/j.disc.2005.06.030. [31] J. Zhang and Y. X. Chang, The spectrum of cyclic BSA $(v, 3, \lambda; \alpha)$ with $\alpha=2, 3$, Journal of Combinatorial Designs, 13 (2005), 313-335.  doi: 10.1002/jcd.20049. [32] J. Zhang and Y. X. Chang, The spectrum of {BSA ($v$, 3, $\lambda$; $\alpha$)} with $\alpha$= 2, 3, Journal of Combinatorial Designs, 15 (2007), 61-76.  doi: 10.1002/jcd.20104. [33] J. Zhang and Y. X. Chang, Existence of BSAs and cyclic BSAs of block size three, Journal of Statistical Planning and Inference, 137 (2007), 1496-1511.  doi: 10.1016/j.jspi.2006.02.009. [34] J. Zhang and Y. X. Chang, Partitionable sets and cyclic BSECs with block size four, Journal of Statistical Planning and Inference, 139 (2009), 1974-1979.  doi: 10.1016/j.jspi.2008.09.007.

Figures(3)

Tables(5)