Network Analysis of Chaotic Dynamics in
Fixed-precision Digital Domain
Chengqing Li
College of Computer Science and School of Automation Science
Electronic Engineering, Hunan University, and Electrical Engineering,
Changsha 410082, Hunan, China Beihang University,
chengqingg@gmail.com Beijing 100083, China
arXiv:1811.04733v2 [nlin.CD] 7 May 2019
Abstract—When implemented in the digital domain with time,
space and value discretized in the binary form, many good
dynamical properties of chaotic systems in continuous domain
may be degraded or even diminish. To measure the dynamic
complexity of a digital chaotic system, the dynamics can be
transformed to the form of a state-mapping network. Then, the
parameters of the network are verified by some typical dynamical
metrics of the original chaotic system in infinite precision, such
as Lyapunov exponent and entropy. This article reviews some
representative works on the network-based analysis of digital
chaotic dynamics and presents a general framework for such
analysis, unveiling some intrinsic relationships between digital
chaos and complex networks. As an example for discussion, the
dynamics of a state-mapping network of the Logistic map in a
fixed-precision computer is analyzed and discussed.
Index Terms—arithmetic domain, chaotic cryptography,
chaotic dynamics, Logistic map, state-mapping network.
I. I NTRODUCTION
Haos is typically defined in the continuous domain, for
C which many interesting dynamical properties hold only
in infinite precision. Yet, implementing a chaotic map in a
finite-precision device is an inevitable process for any of its
real applications or for any simulation in the digital world.
Therefore, the dynamical properties of a digitized chaotic
map have been the main concern in nonlinear science and
chaotic cryptography. When iterating a chaotic map in the
digital domain, its orbit is perturbed by the quantization
errors. Consequently, not a real-valued chaotic orbit but only
a discretized pseudo-orbit is obtained. Such pseudo-orbits are
considered useful due to the well-known shadowing lemma,
which says that every pseudo-orbit obtained via computer
simulation has a true error-free orbit nearby. The shadow-
ing lemma justifies the use of computers to study chaotic
systems and their dynamics. As a consequence, many efforts
were devoted to designing pseudo-random number generators
utilizing chaotic maps [1]. Nevertheless, it was argued [2],
[3] that the dynamics of digital chaotic maps are degraded
to various extents and such dynamical degradation can create
big problems to some applications, for example thwarting
the security of the chaos-based cryptosystems as iterating a
This work was supported by the National Natural Science Foundation of
China (nos. 61772447, 61532020).
Jinhu Lu Guanrong Chen
Department of Electronic Engineering,
City University of Hong Kong,
Hong Kong SAR, China
degraded chaotic map generates a short-period pseudo-random
number sequence with a non-negligible probability [4]–[6].
Despite the theoretic elegancy of the shadowing lemma, it
becomes rather misleading when one looks at the system as
a whole: all shadowing orbits expressed in a particular finite
precision constitute only a negligible part of the whole set
of true orbits. This is not surprising because the states that a
pseudo-orbit can visit are countable, but in continuous domain
they are not. Even worse is that, in a finite-state space, all
pseudo-orbits will eventually become periodic and the period
is bounded by the number of the states therein. Thus, to what
extent the finite-precision orbits can still retain the dynamical
properties of the original chaotic system calls for quantitative
analysis of the set of pseudo-orbits. This turns out to be
a rather complicated problem and cannot be easily handled
by existing theories and techniques. In fact, most existing
work focus on the average behaviors of all pseudo-orbits,
but still cannot tell anything more about the pseudo-orbits
within different fine structures. In [7], some fine structures
in such pseudo-orbits implemented with fixed-point arithmetic
are revealed and discussed.
Recently, complex networks as an emerging research field
can provide useful tools for studying the dynamics of chaotic
time series, first attempted in [8]. Essentially, some earlier
work studying the dynamics of digital chaos can be attributed
to the notion of complex networks [9]. Specifically, in 1986
Binder and Jensen built a state-mapping network (SMN) of
the Logistic map and reported that the counterparts of some
metrics on measuring the dynamics of continuous chaos, such
as Lyapunov exponent, remain to work as usual [10]. In
the following years, Binder reported some more work and
findings: the relationship between a limit cycle of the Logistic
map and its control parameter [11]. Binder and Okamoto
studied the shadowing of unstable periodic orbits (UPO) with
the corresponding limit cycles [12]. In [13], Hsu investigated
the global properties of nonlinear dynamical systems by cell-
to-cell mapping. Kurths and his team presented an analytical
framework for recurrence network analysis of time series [14],
and then further studied the geometry of chaotic dynamics
from the perspective of complex networks [15]. In [16],
Frahm and Shepelyansky analyzed the small-world properties
of Ulam’s networks of the Chirikov standard map and the
Arnold cat map, showing that the number of degrees of
separation grows logarithmically with the network size. In
[17], Shreim et al. studied the dynamics of series generated
by cellular automata, using the parameters of the associate
state-mapping network. The properties of Chebyshev map on
a finite field was investigated with a directed network (graph)
by Gassert [18]. Recently, the functional graph associated with
the iterations of the Chebyshev polynomials over a finite field
was quantitatively analyzed by Qureshi and Panario [19]. It
should be noted, however, that networks typically exhibit self-
organizing fractals (and multifractals) and chaos by themselves
when they are sufficiently large and complex [20].
This paper reviews the existing complex-network-based
dynamic analysis of digital chaotic systems, using the Logistic
map as a typical example. A general framework for analysis
is presented. Furthermore, it discusses how the structure of
the SMN of the Logistic map is changed with the fixed-point
arithmetic precision on a computer. Some fractal phenomena
and power-law in-degree distributions are observed when the
precision is incrementally increased.
The rest of the paper is organized as follows. Section II
briefly reviews the existing dynamic analysis of digital chaotic
systems from the general viewpoint of complex networks.
Some special properties of the SMN of the Logistic map are
reported in Sec. III. The last section concludes the paper.
Figure 2 depicts the largest in-degree of the corresponding
space network under different control parameters with five
network sizes. From Fig. 2, one can see that the largest in-
degree obviously follows a power-law distribution with the
same scaling exponent for N ∈ {2000, 10000}. In contract,
there are two scaling exponents for N ∈ {1000, 5000}. This
may be caused by the non-uniform distributions of the floating-
point numbers in different sub-spaces.
Figure 3 presents the average path length of the space
network of the Logistic map under different control parameters
with five network sizes. Observing Fig. 3, one can see that the
changing trend of the average path length with respect to the
control parameter matches that of the Lyapunov exponents of
the Logistic map to a relatively high extent, demonstrating that
the parameters of the phase network can serve as an alterna-
tive measure for detecting the ‘edge of chaos’ in dynamical
systems.
103
102

II. B RIEF SURVEY ON COMPLEX - NETWORK - BASED 101

10-3 10-2 10-1 100
DYNAMIC ANALYSIS OF DIGITAL CHAOTIC MAPS

Following Ulam’s method proposed in [21, p. 74], the state- Fig. 2. The largest in-degree kmax of the space network of the Logistic map
mapping network of the Logistic map in five network sizes.

f (x) = µ · x · (1 − x) (1)
50
N = 2048
N = 1024
was built by the following three steps in [22], [23]: 1) divide 40 N = 512
N = 256
the phase interval into some subintervals; 2) each subinterval N = 128
30 Lyapunov exponents
is considered as a node; 3) two nodes are linked by a
directed edge if a (predefined) map relationship between the
davg

20

corresponding intervals exist.

10
Once the network is built, various parameters can be calcu-
lated to reveal some rules similar to that obtained in the infinite 0

resolution domain. The general framework of analyzing the -10

dynamical properties of chaotic systems can be shown by the 0 0.5 1 1.5 2
µ
2.5 3 3.5 4

four stages in Fig. 1.

Fig. 3. The average path length of the space network of the Logistic map
with respect to the control parameter µ for five network sizes.
Establishing link Defining network Comparing with other
Selecting nodes
between nodes parameters recognized metrics
In [24], the time series generated by a chaotic system is
Fig. 1. Framework of network-based analysis of dynamical systems. transformed to a network (graph) using the mechanism of
horizontal visibility graph, where every state is considered as a
In [23], the phase space of a chaotic map g(x) is equally node and two arbitrary data points are linked if and only if they
divided into N disjoint sub-space, and each sub-space is satisfy the geometric ‘visibility’ criterion. Then, a parameter
defined as a node. Given an initial state, x0 , the ith node is called “network entropy” is defined. As shown in Fig. 4, the
directly linked to the jth node if and only if x0 and g(x0 ) changing trend of the network entropy of the associate network
belong to the two corresponding sub-spaces, respectively. of the Logistic map can mimic that of the Lyapunov exponent.
Iterating the chaotic system from the current state T several In [25], two arbitrary states of a time series are connected as a
times and linking the related nodes in a similar manner, the pair of nodes if the linear interpolation between them is larger
space network is established. than the values of all intermediate points. The distribution of
 2
h more complex ones found in CAs falling into Wolfram’s
log 4 classes III and IV (see the in-degree distribution for four
1
rules shown in Fig. 6). Yet, the validity of such classification
λ method was questioned in [26], which plots the distribution
0 of the two parameters corresponding to all elementary rules
of 1-D CA falling into the same Wolfram’s class. How to
-1 accurately measure the dynamical complexity of a chaotic
system with the methodology of complex networks needs
further investigation.
3.6 3.7 3.8 3.9 4
µ It has been observed that, in the field of nonlinear science,
many did not care about how their computers perform fixed-
Fig. 4. Dynamical demonstration of the Logistic map under two metrics:
horizontal visibility network entropy h and Lyapunov exponent λ. point or floating-point arithmetic operations, but implemented
a chaotic system in the computer just like a black box.
Most related research on dynamical analysis based on network
different subgraphs is shown to be a sensitive measure of the (graph) adopt the Logistic map as test objects [1], [2], [10],
underlying dynamics. As shown in Fig. 5, motif distribution [14], [15], [22]–[25], where the real structure of SMN of the
can be used to distinguish broad classes of chaotic dynamical Logistic map in the computer has been unfortunately ignored.
systems, to which the Logistic map belongs.
III. DYNAMICAL ANALYSIS OF A STATE - MAPPING
NETWORK OF THE L OGISTIC MAP ON COMPUTER

4
Now, consider the following state-mapping network of the
10

Logistic map
Logistic map: every value in the n-bit arithmetic computing
Henon map
Ikeda map
precision is identified as a node; a directed edge between nodes
Cat map
Tinkerbell map
u and v exists if and only if the former node is mapped to the
3
10 Standard map
Duffing map
latter one by the Logistic map. The following key parameters
Tent map
of the space network were studied, revealing the relationships
among the Logistic map (1), the computing precision and the
10
2

A B C F D E
control parameter.
• Loop: (also called a self-loop) is an edge that connects a
Fig. 5. Subgraph ranks of different types of time series, where A, B, C, D,
E, and F represent 6 different possible undirected subgraphs of size 4. node to itself.
• A directed cycle in a directed graph is a sequence of
nodes, starting and ending at the same node such that,
0
Rule 4 Rule 22 Rule 110 Rule 184 for each pair of two consecutive nodes in the cycle, there
-0.1

-0.2
L=15
L=20
exists an edge directed from the previous node to the later
-0.3
L=25 one.
A weakly connected component (WCC) is a maximal
logP (k)/logN

-0.4 •
-0.5 subgraph of a directed graph such that there is a path
-0.6
between every pair of nodes u and v in the subgraph,
-0.7
regardless of the direction of edges.
-0.8

-0.9 To observe the properties of SMN of the Logistic map,

-1
0 0.5 1 1.5
logk/logN
Fig. 6. In-degree distribution P (k) for four rules of CA with three values of
CA having size L, where N = 2L .
In [17], the state-space network of the sequential dynamical
systems is analyzed, which is composed by 1-D cellular
automata (CA) with rules involving two colors and the nearest
neighbors, where every binary state of size L is viewed
as a node of a directed network and the evolution relation
between two nodes builds a link connecting them. Based on
analytic analyses and experimental results on some selected
rules, it was claimed that the co-appearance of nontrivial
scalings in both the hub sizes and the path diversity of the
state-phase networks can separate simple dynamics from the
2
the software platform Cytoscape is used to draw some space
networks under some given parameter values. The space
networks of the Logistic map of µ = 121 25 under 5-bit and
6-bit fixed-point arithmetic are shown in Fig. 7a) and b),
respectively, where a round quantization is adopted. To show
the rule of the network with respect to the increase of the
precision more clearly, the state network of the Logistic map
under 6-bit precision with the same value of the control
parameter is shown in Fig. 8. When the precision is further
increased to 12, the network becomes the case shown in Fig. 9,
demonstrating some fractal phenomena, as studied in [27]. As
shown in Fig. 10, the in-degree distribution of Fn approaches
a power-law more steadily as n increases, where Fn denotes
the function of Logistic map (1) implemented under n-bit
precision.
 TABLE I
I NTERMEDIATE VALUES OF THE CALCULATING L OGISTIC MAP IN BINARY 16.

x 1−x 1 − (1 − x) f (x) f (1 − x)
0.0099945068359375 0.98974609375 0.01025390625 0.037384033203125 0.038360595703125
0.04998779296875 0.94970703125 0.05029296875 0.179443359375 0.1805419921875
0.0899658203125 0.90966796875 0.09033203125 0.309326171875 0.310546875
0.0999755859375 0.89990234375 0.10009765625 0.340087890625 0.340576171875
0.199951171875 0.7998046875 0.2001953125 0.6044921875 0.60498046875
0.289794921875 0.7099609375 0.2900390625 0.77783203125 0.7783203125
0.389892578125 0.60986328125 0.39013671875 0.89892578125 0.8994140625
0.489990234375 0.509765625 0.490234375 0.9443359375 0.94482421875

0
37 22 42 12 19
10
5 18 15 12 20 3 38
3 56 39
61 52
29
31 26 60
11 11 27 −1
16 14
28 53
2
25
15
45 10
17 35
27 9 36 8
10 49
4 21 4
62 9
30 1
33 −2
10
17
10 55 63
22 25 31 47
1 7 54
28
2 34 29 21
43 48
32
0 30 23 58
26 20 44 50 16 −3
14 6
10
19 46 51 18 5 41 23
32 6 13 8 24 57

13 40 59 24 7 64 0
−4
10
a) b)
−5
10
121
Fig. 7. The state network of the Logistic map with µ = 25
under round
quantization: a) 5-bit precision; b) 6-bit precision. 10
−6
0 1 2 3
10 10 10 10
k +1

2 126
27 31 77
45

83
92 97 78
8 120
51 Fig. 10. In-degree distribution of the sate networks of F5 ∼ F20 .
95 9
114 101 50
35
119 29
32
33 93
49 36 96 99
118 10 34 94 88
14 23
17

79
98
84
1
127
111

107
40

74
105
weakly connected components;
90 44 58 21

25 30
104
113
4

124
69
22
122
54

48
• Each weakly connected component has one and only one
103 24 71 6

91 37
38
76 52
15

57 59
70 106 117 80 self-loop or cycle;
75 89
39

53
121

20
68
11
3

61
• In each cycle of length longer than 2, there is at least one
82
7 26

47
108
66
67

125
62

64
node with a large in-degree;
Let (Dn /2n ) denote the node with the largest degree in
115 81 12
86 18
102
116
60 63
65
41
•
46 0 110 109
28 43 42 73

13 128 100 85 112

16

55
56
123 72
19

5 87 the state network of Fn (x). Then,

Dn+1 ∈ {2 · Dn , 2 · Dn + 1, 2 · Dn − 1}, (2)
Fig. 8. The state network of the Logistic map under 6-bit precision and round
quantization, where µ = 484 . Nµ
27 where n ≥ nµ , µ = , Nµ , nµ ∈ Z+ .
2nµ
• One weakly-connected component dominates the whole
1056

1132
2964
3040

3097

1240
284

network, i.e., the size of the component accounts for more

2856
827
999
3269

1793
2303
1600
2496
744

1794 2918

3812
2302 3687
409 1178

2304
3173

3352 1792 1392

491
1510

382 547
2704
2786 3549
727
2262 923
3605
1634
3475

2586 3369 1834

1310 2261 266
3830

3714
3156
812 621 171
1992 3925
940
1835
2462 1357

2739 620
2703
1250

3284 665

3431

3476
1990

994
2846 2106
1393

3102 3174
2961 922

2702

1135

88 1716
326 3879
3477 1394

1715

3770 777 1987

4008

2381 217

619

2380

3319

2822

3433

1274

2101

3869
663

666

2107
2109
2740

3430

2103

1356
397

1994

2521 3699

2481
1574 1615 1991

3665
2522
431

1358
3204
1988

1575
1458
3432 2104 4035

2638

892 2738
3551

208
664 227
61

than half of the size of the whole network. Among all the
1787 184 3912

3888 747 545 1989

1995

622 811
1993
3809
2309 2102
2108 2105
1786
3474
1295 3285
3349

2310 287 2705

2970 2459

1391

2801 1009

3087
1126
971

3125

2875

1221
3241

3993
1542

853
125 1555 103
2558 855 2541
3971
460 380
2554 3636
458 3648

792 1530
3243 1544 122

140 3308
1545 3304
788 2517
1509
448
2416 427
1538 875 3624
3638 507 349 2406 2566 3716
3956
511 492
852 464 3669 3974 2792
2551 3632 1579
2650 3010
1208 2358
2552 3747 1690
1446 1304
1680
2602 192
3249 1207 1086
1490 3754 1737
3221 2760
877 1691 313
3589 472
129 1159
1556 342 3537 1637 735 3018
2873 3244 2937 3783
2911 559 3361
3585 2405 3904
2540 912 692 2359
2888 3604
3404
3142 1330 1093
1494 2606 1336
3649 847 3967 1825 2281
1415 270
1185 1097 2280 3003
443 1738
1223 3184 1059 21
2175 1824 2766 1078
594
22 96 2681
1557 1329 3030
3219 3396 3822
354 2539 3653
4013 3037 3826 1107
3502 954
1815 2393 1292
3742 4075
3301 2969 2767 3050 1046
700 79
1127 731
2535 701 1070 4017
2176 2999
83 2423 1920 2272 1703 752
2271 3395 2195 1041 2982 1066
1561 1816
4000
4074 307 244 274 2945 3762
3089 3055 3344 293
1673 2989
3022
447 965
1922 2197 2194 2804
795 2321
1100 2936
3365 3067
1506 1007 1774
1543 967 1114
763 2789 1900 3026
3129 1151 1160
1054 2872
1074 3857
2996 2587
1919 868 334
1749 2590
87 3803 1029
1284 3144
3042 2553 1307
239
3131
3601 3637 1636
1302 1152
2348 1136 541
3333 2960 2974 3245 1224 459 1775 2322

1049
3789 381
322 3852 2793 2763 952
925 2347 495 4009 1901
2889 851 1541
2944
1388 3104 1646 3047 3715 2461 4041 193
992 2794 3358 738 1 3903 2913
2812
1480 1635 1303
3080
1921 55 3555
2951 2196
1748 2616 1122 1462 1183
2708 3400
2288 1624 2762 695
2843 3635 2555 205
1333
2174
3470 1145 4092 2177 2549 2460 3360
3437 531 3574 586
3171 1253 277 4095 4
626 659 693 3228 2601 2634
152 1808 3181
2736 3119
4096 0 3565 2185
1037 1723
3913 1899 3706 3401 3206
4047 2625 3289 15 1547 2675
2005 2091 3819 143 2182
915
909 736 696
1168 1360 3403
1902 1013
2098 1722 2883 3891 2472 1334
977 461 890
553 3509 1233
3590 2687
3435 2095 522
1471 1809 2373 1213 1495
2928 2287 4081
183 3774
661 2092 3944
2813 1283 1335 2674
2184
57 2761 3145 3510 2863 3855
3059 3157 13 3745 1677
587 1409
390
49 2089 1421 3083
3259
1517 939 3953
2096 3495
3543 3640 3187 351 106
3332 1914
2006 2374
108 506 2283 837
1451 213 2844 951
484 2737 2447 95 4045
4039 1016 1252 2419
3988 3880 3990
2518 1911 601
4083 2002
3612 774 456 114 1422 1912 51
4001 591
2645 1939
3322
2579 78 1783
398 764 289 3287 419
1614 3725
1999 373 3201 1532 2183 1913 241 190
997 1276
3883 3668
216 1649 1782 3677 1578
1244 895 3505
2820 858
1996 2564 2878
3807 3982 3821
1133 4018 2455 2159
2450 58 809 685
2852 226 2814 371 2508 1449
3723
3906
2350
3099 4012 4038 3626 1588
2314 2156
3275 2284 3861
84 2313 3719 1814 2157
428
766 3236 3541 1531
3347 1219
1282 1294
470

connected components, there is a clearly decreasing order.

3621
1218 1941
311 3148
1641 475 2569 1813
2963 662
807
3788
1010
1741 1573
749 1060 3411
3330
3698 1359 2349 555 2410 2565
1938
2282 2755
3015 3238
1747 377 1812 471
2802
1081 1527 2647
2482 2877 3625
2355 860
2158
3434
230 1746 2971
1937 2155 1341
3751 1940 948
3009
2099 3200 3086 1576
3785 3036
896 1686
821 3120 976 1195
1077
1969 308
673 2126 838
1965
2130 4033
1125
2811 2901 323
374 235
345 2810 1286
1087
2520
3873
2127 223
3423 1967 2627
1998
761 2598
1280 3336
1285 1721
3335 2836
1720 63
797 212 3562
1970 1645 3258
3722
612 1469 366 1498 2816
2001 3019 2343 3563 3773
3593
3328
760 2341
2128 3299

1260 2427 768 1756

3866
1968 1753 2375 2376 3321
2451 534
3792
2004 1668 3884 3793
432 2360
2131
839
1400 2635 3739
3484 542 3730 527
533
1656
2342 503
56 67
3093 775 3666 1475 2340 1755
1754 285 1735
4040 1076
2007 3020
1669 3782
3485 2428 3212
1966 1736
2696 2440
4014 209 3569
1094 249 314
3000
674 303
1293 2621 1061
2129 4029 2361
2090 972 357
2673
3002 1064 304
611 3887 1232 3979 3935

750 884 43
2306
1096

2746 3422 2803

3870 82 3346 1423
2094 3035
430 1324
3511
161 585
1148 2317
3124 1455 708 3901
117
1350 1791
395 3032 4053 1896
2097 3388 2211
1779 1882
1123
2318 956
2948 2772 2214 2973
702
195

3547 1884
2641
2100
2200
3851
3140 1014 2199
2771 2213 3082
2669
707
4022 1798
549 3394
1898
2885 1211
3158 2198
9
3991
1790 275 3516
2212 1897
291 745 970 3814 1427
2298 3389 3847
3225
485 1618 2299 1883
74 28
1516 4062 2735
2479 34 938 871
4087 387
386
2992 1797 105 245
865 2774 1227
3436 1325 580
1617 127 2093
3701
3611
1215 4020
2869
969 2580 1629 1322 993
1104
4068
1124 711
3049
3969 1361 3385 3250
3710 282
3231 2478
465 660
1196 2468 3709
2799 2593
1299 846
3031
2797 3084 2559 2222 2947 2845
3127 2221 1885
2881 2978 76 2538
3277 1118
819 2467 3126
252 1065
2972 1559
3354 3631 1047
1873 3103 1628 160
581
1245 4050 489 2980
3515
1012 2851 624 3650
742 288 3651 3844 1149
1558
3735 1537 894
1454 2297 577
1500 1390 2000 3074 3607
722 261 2670
1663 172 3936 1426 3070
361 3595 2537 363 1503
1997 2542
1874 1875
2311 46 1116
2247 623
1785 3189 2908
3472 1512
2433 445 1848 378 2584 446
3546 1799 2223 867
3815 1508 3647
2959 2588 907
281 2249 924
501 2596 1430 3519
2456
2642 1800 3808 3374 3473
1639 1026
2637 3924
3835
3603 1188
3959 449 1554
748 2706 3229
2296 1849 3598
990 1137 2312
3106 1784 137
3552 3529 2666
550
1847
343
3753
2248 2407
151 3997 493
905 1640
2839
3348 1459
544 3191
3945
567 2907 498
1225 1689
810
365

1654
3294 1257 99 156
1177 2871 1189
3940
3172 2840
2003 3247

2439 3731

2594 849
3729
625
3064 3293
2919
802 1657 3471 2545 1453 2442 2441
3597
1256
551
1655
2707
1032 1389 3356
1502 3644
499 294 1551 803

2494

206
1521 2293

408 740
452
1601 3688 3616

111 367
1602
1263

3890
3985 2575
2957
2495

1804

In the floating-point arithmetic domain, the symmetric prop-

480

3109
4061
870 1805 1139
131 3303
3226
2833
3965
2291

2591
2292
540 35 987
2418

1772 2842
806 1505
3600 793
1463
2323 1679
3746
1773 364

3556
3802 3290
2448 1648 2417
1254

1678
1658 496
2633 3724

2324
4069

1647
3732
2437 1659 2449
27

101

210
1210
753
2438
350

372
3224

3886 3995
3343 2886
1279
2595

3596 872

279

1660
3117 1501

500
3326 2817

1642
3952 3295
4080
979 1650
144
1477 1291 805 370
883 3720
2805
2384 328 2955 16

525 778
1712 60
2364 3726
2454 2446
1203 2619 3881
3571 376 1255 1273
1733 1141
3768 3113
3213 770 3291
1731 3780
2383 224

2366 3994
102
1713
4036

3779 3318
2893 1103 215
1732 2363 2841
134
1018 3123

1730 3048 991 2071

2585
316
2365
617 3962
3479 3078 3105 982 2069
185 2949
490
3428 317 276 973 2823

3606 2993
1982 3024
2115 1106
2112 1511
668 1142
1631
1986 3282 3114
2114
3911 228 1810
1983 2990 2286
1072
3053
1980 3045
1147
2466 3711 2954
1630 3820
814
667 737 330

1355 2285
2940 1043
2646
3766
2465 1811 81 1051 1271

3542 2387
3429 2111
300
941
1981 3359
385 3137
1156 1709
2741
554
1048
1985 1450
1212
4015
2777 1761
2334
1319 299
3868
3155 2110 959 2865
2884
3797
1167 3227 3796
2113 1231
2335

2929 869 3256

1763
2117 1433
618 2589 1762
2333
2663
3381
1984 2525
3560
3523
3339 536 2067
2742 715 840
757 3602 1467 3662
2701 1507
2452 147
3478 211
573 2066
2116 1288
3427 3938
2232 4060 2629
1864
1643
2808 3949 1571 2064
3885 3721 434

1395
410 1354 2062
158 1644 494 135
669 375 781
1979 574

erty of the Logistic map does not hold in general [3]. To

3926 2453 36
3686 616
3480
1396 3961

2497 808
3288

1599 170
2574
2700

3617

479

2061
1522

1706

2059

2058

1865

2230

1866

3841
2231 332

3522 1432

2942
904

2664

2056
962

2054
1633

3713
2463

1313
255 2053

2051

383
104

20492048

1262
3992
2834
2046

3302 3895

3345

3192 222
723
2421
2420

1777 794
1190 2319 54 201
2906 751
3744
1418

1025 3804 1675

3874
2045
3071 4042
590 1676
3506 2043
1776 2320 2678 352
292

65 1217
1910
3085
2189 4031

2186
589 242
2879
3235
2677

1908
1011
1907
3507
3854
418
861
1320
1419 1528 2041

474
566
1909
713 2040
1439

1589 1667 3195

3383
3622
254 2038
2188 2571
2776 1844
1585

1590 2429 2227 2228 3530

2568 2523
2187 3664 3257 2036
2657 1461
3678 2864
3738 1867 3554
901 1003
2506 2572
697 3619 1778 2035
2861 422 3811
2665 575
3805
3399

demonstrate this point, some intermediate data obtained from

2305
2507 2764 1140 1235 3842 1022
97
1525 1524 1868
985 4037 3351
3262 1869
358 2900
3264 59
1052
1332 3521 1297
832 1267
3111 2229 3202
220 3382
2956 1431
2829 3718
477 3999
2643

3309 3674 2457

1594
3044 2033
787 714
1101 3545
3876
3286 2032

1803

2404 2030
2995
1251

2511 3817 2028

1692 3733
2027
3755
983

2025 29
2436
234
798
3298 341 2825
1182
3987
834 801
109
2430
3315 4067
2486

502 3180 2914 1845 3872 401

1665
359
1846 2390
3737
3695
1666 414 2250 2063
2689
1368 1610
916 2252 3764
1499
1238
3594
2431 262 3448
830

3682 1154

2726 3161
3445
2597
1407 3134
2728
2502
873 934
961

3223 3266 2783 2070

3832 935

3373 3135
2781
1839 3916 1315
651
1000
2251
835
1057 3532 3376

2858 3834
1098 1342 424
180 2659
2655
3261
2754 569
927
3493 2514
2965 3196 2727 1583 720
2244
3039 603
2712
3412
3096 1192 1437
1075 3527
3465
2998 3752 3672 1853
1369
2018 473
3446
684 2409 2242
3871
2154 3307 2513 3837
804 1582
2443 1266
1944 3728
859
3021 2068
1852
2151

1088 1131
325 2567 3623
650 259
1946
3008 2243
1854

1943
3237
3169 3447
917
3179
1384

1529
631
1718 2065
776 2691
1548
2379 3641
1405
4063
649 455
3929
1387

823 124
1275 2378 3771 2550 3490
1546
2008 2548
3320 1945 3469
3972
606
3639 3273 33 167
1717 2487
2150
2709 1609
3907
2821 627
1952 189

520 2153
981 3576 3862 3694
1535
457 3923 682
173
2011
3115
1482 3930 628
3629 2752
2614 402
3657 1371
1565
2149
2561 3414
3468
602
439

166 3449
2531 467 1942
629
2725 680 1344
2267 2013

729
2060
3494 1345
407
2690