A Time-varying Traffic Sharing Protection Based on Spectrum Window Sliding in Elastic Optical Networks
-
摘要: 针对弹性光网络(EONs)中时变业务生存性传输时消耗频谱资源多和业务阻塞率高的问题，该文提出一种基于频谱窗滑动的时变业务共享保护(TTSP-SWS)算法。TTSP-SWS算法选择可用频谱块承载权重和保护频谱块共享度的共享保护路径代价函数值最小的保护路径；通过在保护路径上滑动频谱窗，将时变业务分配至频谱窗共享度最高的频隙位置；当时变业务带宽变化时，采用基于频谱窗滑动的生存性扩展或压缩频谱分配策略调整时变业务的带宽分配。仿真结果表明，该文所提TTSP-SWS算法能降低网络的业务阻塞率和保护资源冗余度。Abstract: Considering the problems of high spectrum resource consumption and high traffic blocking probability during the time-varying traffic’s survivability transmission in Elastic Optical Networks(EONs), a Time-varying Traffic Shared Protection method based on Spectrum Window Sliding (TTSP-SWS) is proposed in this paper. In TTSP-SWS, the protection lightpath with the minimum cost including the load weight of available spectrum block and the sharing degree of protected spectrum block is selected as the protection lightpath for time-varying traffic. By sliding the spectrum window on the protection path, the spectrum window with high sharing degree is allocated to time-varying traffic. When the required bandwidth of time-varying traffic changes, a spectrum window sliding-based survivable spectrum expansion or compressed strategy is adopted to adjust the allocated bandwidth. Simulation results show that TTSP-SWS can reduce traffic blocking ratio and protection resources’ redundancy.
-
算法1 基于频谱窗滑动的频谱共享生存性RSA策略 输入：弹性光网络拓扑G(V, E, F)，时变业务r(s, d, B, q)，工作路径$ P_{{\text{sd}}}^w $以及频谱分配情况； 输出：时变业务r(s, d, B, q)的保护路径$ P_{{\text{sd}}}^p $及其频谱分配结果； (1) 将时变业务的工作路径$ P_{{\text{sd}}}^w $从网络拓扑G(V, E, F)中移除； (2) 执行K最短路径算法，根据式(9)，计算每条候选保护路径的共享保护代价值$ C(P_{{\text{sd}}}^p) $，并升序排序候选保护路径在集合
$ P_{{\text{sd}}}^{{\text{ps}}} = \{ P_{{\text{sd}}}^j,j = 1,2,\cdots,K\} $中；(3) 根据各候选保护路径的长度，计算确定业务所需最高调制等级，并由式(1)计算业务在各候选保护路径中所需FS数目f； (4) 依次统计各候选保护路径中满足时变业务所需FS的可用频谱块，并保存在业务的可用频谱块集合Bj, A中； (5) 若Bj, A集合非空，转步骤(6)；否则，阻塞时变业务r(s, d, B, q)，结束算法； (6) 在非空的Bj, A集合中，选择路径序号j最小的可用频谱块的候选保护路径$ P_{{\text{sd}}}^j $标识为$ P_{{\text{sd}}}^p $，将$ P_{{\text{sd}}}^p $中可用频谱块的共享FS保存在集合
BP中，若BP中存在满足时变业务r(s, d, B, q)所需FS数目的频谱块；转至步骤(7)，否则，转至步骤(11)；(7) 从BP中选择包含共享FS数目最多的共享频谱块，标识为BP,m，将共享频谱块BP,m的起始索引值记为index(FP)，末位索引值记为
index(EP)；(8) 在该保护路径上创建业务的滑动频谱窗SW, SW的大小由时变业务r(s, d, B, q)所需FS数目f决定； (9) 根据式(10)，依次计算${\text{S} }{ {\text{W} }_{ {\text{index} }({F_{\text{P} } })} }{\text{～}}{\text{S} }{ {\text{W} }_{ {\text{index} }({E_{\text{P} } } - {f_{} } + 1)} }$的频谱窗共享度$ W_m^s $值，选取$ W_m^s $值最大的SW，并将时变业务r(s, d, B, q)所需
保护频隙数目分配在频隙索引值区间$ [m,m + f - 1] $，输出结果，结束算法1；如果存在$ W_z^s = W_n^s $, $ z,n \in m $，并且$ z \ne n $，执行下一
步骤；(10) 计算$ d_{\text{L}}^{\text{P}} = \varepsilon - {\text{index}}({F_{\text{P}}}),d_{\text{R}}^{\text{P}} = {\text{index}}({E_{\text{P}}}) - (\varepsilon + f - 1) $，$ {d_\varepsilon } = |d_{\text{R}}^{\text{P}} - d_{\text{L}}^{\text{P}}| $，其中$ \varepsilon \in (z,n) $；选取$ {d_\varepsilon } $值最小的频谱窗，并将时变业务
r(s, d, B, q)所需保护频隙数目分配在频隙索引值区间$ [\varepsilon ,\varepsilon + f - 1] $，输出结果，结束算法1；(11) 从保护路径$ P_{{\text{sd}}}^p $的Bj, A中选择包含共享FS最多的可用频谱块BA, m作为预分配频谱块，该频谱块包含频隙数目为$ |{\text{B}}{{\text{P}}_{{\text{max}}}}| $，并确定BA, m
的起始索引值index(F)和末位索引值index(E)，并确定该可用频谱块内的最大共享频谱块的起始索引值index(FP)和末位索引值
index(EP)；(12) 计算$ d_{\text{L}}^{\text{A}} = {\text{index}}({F_{\text{P}}}) - {\text{index}}(F) $, $ d_{\text{R}}^{\text{A}} = {\text{index}}(E) - {\text{index}}({E_{\text{P}}}) $，并计算$\varDelta = {f_{} } - |{\text{B} }{ {\text{P} }_{ {\text{max} } } }|$； (13) 若$ d_{\text{L}}^{\text{A}} > d_{\text{R}}^{\text{A}} $，则更新$ {\text{index}}({F_{\text{P}}}) = {\text{index}}({F_{\text{P}}}) - 1 $； (14) 若$d_{\text{L} }^{\text{A} } \le d_{\text{R} }^{\text{A} }$，则更新$ {\text{index}}({E_{\text{P}}}) = {\text{index}}({E_{\text{P}}}) + 1 $； (15) 更新$\varDelta = \varDelta - 1$； (16) 若$\varDelta \ne 0$，返回步骤(12)；否则，将时变业务r(s, d, B, q)所需频隙数目分配在频隙索引值区间$ [{\text{index}}({F_{\text{P}}}),{\text{index}}({E_{\text{P}}})] $，输出结果，结
束算法1。
下载: 导出CSV
算法2 基于频谱窗滑动的生存性扩展频谱分配策略 输入：时变业务r(s, d, B, q)，带宽扩展后的时变业务r(s, d, Bn, q)，当前保护路径的FS状态； 输出：带宽变化后时变业务r(s, d, Bn, q)的频谱分配结果； (1) 根据式(1)，计算带宽发生变化后时变业务r(s, d, Bn, q)在保护路径上所需保护FS的数目$ {f^n} $； (2) 若时变业务r(s, d, Bn, q)当前占用保护路径的可用频谱块满足$ {f^n} $需求，即$ {B_{{\text{A}},{\text{ }}m}} \ge f_{}^n $，其中，BA, m表示当前时变业务r(s, d, Bn, q)
在保护路径上占用可用频谱块的大小，并将该频谱块保存在集合EA, m中；否则，从业务的保护路径集合$ P_{{\text{sd}}}^{{\text{ps}}} $的可用频谱块集合Bj, A中
寻找满足$ {f^n} $需求的其他可用频谱块，并确定对应的保护路径$ P_{{\text{sd}}}^j $，可用频谱块保存在集合EA, m中；若EA, m非空，转步骤(3)，否则，阻
塞业务；(3) 若EA, m的保护频谱块能满足$ {f^n} $需求，即$ {E_{{\text{P}},{\text{ }}m}} \ge f_{}^n $，则将该保护频谱块保存在集合EP, m中，转步骤(4)；转步骤(7)； (4) 从EP, m中选择包含共享FS数目最多的共享频谱块，记录该共享频谱块的起始索引值记为index(FP)，末位索引值记为index(EP)，根据
时变业务r(s, d, Bn, q)带宽变化所需FS数目$ {f^n} $，创建滑动频谱窗SW；(5) 根据式(10)，依次计算${\text{S} }{ {\text{W} }_{ {\text{index} }({F_{\text{P} } })} }{\text{～}}{\text{S} }{ {\text{W} }_{ {\text{index} }({E_{\text{P} } } - {f^n} + 1)} }$的频谱窗共享度$ W_m^s $值，选取$ W_m^s $值最大的SW，并将时变业务r(s, d, Bn, q)所
需保护频隙数目分配在频隙索引值区间$ [m,m + {f^n} - 1] $，输出结果，结束算法2；如果存在$ W_z^s = W_n^s $，$ z,n \in m $，并且$ z \ne n $，转步
骤(6)；(6) 计算$ d_{\text{L}}^{\text{P}} = \varepsilon - {\text{index}}({F_{\text{P}}}){,_{}}d_{\text{R}}^{\text{P}} = {\text{index}}({E_{\text{P}}}) - (\varepsilon + {f^n} - 1) $, $ {d_\varepsilon } = |d_{\text{R}}^{\text{P}} - d_{\text{L}}^{\text{P}}| $，其中$ \varepsilon \in (z,n) $；选取$ {d_\varepsilon } $值最小的频谱窗，并将时变业务
r(s, d, Bn, q)所需保护频隙数目分配在频隙索引值区间$ [\varepsilon ,\varepsilon + {f_{}} - 1] $，输出结果，结束算法2；(7) 从保护路径的可用频谱块保存在集合EA, m中选择包含共享FS最多的可用频谱块作为预分配频谱块，其大小记为BE，确定该可用频谱
块的起始索引值index(F)和末位索引值index(E)，并确定该可用频谱块内的最大共享频谱块的起始索引值index(FP)和末位索引值
index(EP)；(8) 计算$ d_{\text{L}}^{\text{E}} = {\text{index}}({F_{\text{P}}}) - {\text{index}}(F) $，$ d_{\text{R}}^{\text{E}} = {\text{index}}(E) - {\text{index}}({E_{\text{P}}}) $，并计算${\varDelta _{\text{E} } } = f_{}^n - {B_{\text{E} } }$； (9) 如果$ d_{\text{L}}^{\text{E}} > d_{\text{R}}^{\text{E}} $，则更新$ {\text{index}}({F_{\text{P}}}) = {\text{index}}({F_{\text{P}}}) + 1 $； (10) 如果$ d_{\text{L}}^{\text{E}} \le d_{\text{R}}^{\text{E}} $，则更新$ {\text{index}}({E_{\text{P}}}) = {\text{index}}({E_{\text{P}}}) + 1 $； (11) 更新${\varDelta _{\text{E} } } = {\varDelta _{\text{E} } } - 1$； (12) 如果${\varDelta _{\text{E} } } \ne 0$，转至步骤(8)；否则，将时变业务r(s, d, Bn, q)所需保护频隙数目$ f_{}^n $分配在频隙索引值$ [{\text{index}}({F_{\text{P}}}),{\text{index}}({E_{\text{P}}})] $区间，输
出结果，结束算法2。
下载: 导出CSV
算法3 基于频谱窗滑动的生存性压缩频谱分配策略 输入：时变业务r(s, d, B, q)，带宽压缩后的时变业务r(s, d, Bw, q)，当前保护路径资源分配结果； 输出：带宽压缩后时变业务r(s, d, Bw, q)的频谱分配结果； (1) 根据式(1)，计算带宽压缩后时变业务r(s, d, Bw, q)在保护路径上所需保护FS的数目$ {f^c} $，将时变业务在当前保护路径分配频谱块的起始
索引值index(F)和末位索引值index(E)；(2) 计算时变业务带宽压缩后，在保护路径上需要减少的FS压缩$ {\varDelta _{\text{C}}} = {f_{}} - f_{}^c $； (3) 计算时变业务r(s, d, B, q)在当前保护路径占用频谱块的起始FS的共享度$ {W_{{\text{index}}(F)}} $；若$ {W_{{\text{index}}(F)}} = 1 $，转步骤(4)；否则，转步骤(7)； (4) 更新频谱块的起始频隙索引值，$ {\text{index}}(F) = {\text{index}}(F) + 1 $； (5) 更新$ {\varDelta _{\text{C}}} = {\varDelta _{\text{C}}} - 1 $； (6) 若$ {\varDelta _{\text{C}}} = 0 $，将时变业务带宽压缩后所需保护频隙数目$ {f^c} $分配在频隙索引值区间$ [{\text{index}}(F),{\text{index}}(E)] $的频谱块上；若$ {\varDelta _{\text{C}}} \ne 0 $，转至步
骤(3)；(7) 计算时变业务在当前保护路径占用频谱块的截止FS的共享度$ {W_{{\text{index}}(E)}} $；若$ {W_{{\text{index}}(E)}} = 1 $，转步骤(8)；否则，转步骤(11)； (8) 更新$ {\text{index}}(E) = {\text{index}}(E) - 1 $； (9) 更新$ {\varDelta _{\text{C}}} = {\varDelta _{\text{C}}} - 1 $； (10) 若$ {\varDelta _{\text{C}}} = 0 $，将时变业务带宽压缩后所需保护频隙数目$ {f^c} $分配在频隙索引值区间$ [{\text{index}}(F),{\text{index}}(E)] $的频谱块上，输出结果，结束
表3算法；若$ {\varDelta _{\text{C}}} \ne 0 $，转至步骤(7)；(11) 根据时变业务r(s, d, Bw, q)带宽压缩后所需FS数目$ {f^c} $创建滑动频谱窗SW； (12) 根据式(10)，依次计算${\text{S} }{ {\text{W} }_{ {\text{index} }(F)} }{\text{～}}{\text{S} }{ {\text{W} }_{ {\text{index} }(E - f_{}^c + 1)} }$的频谱共享度$ W_m^s $值，选取$ W_m^s $值最大的SW，将时变业务带宽压缩后所需保护
频隙数目$ {f^c} $分配在频隙索引值区间$ [m,m + f_{}^c - 1] $的频谱块上，输出结果，结束算法3；如果存在$ W_z^s = W_n^s $, $ z,n \in m $并且$ z \ne n $，
转至步骤(13)；(13) 计算$ d_{\text{L}}^{\text{C}} = \varepsilon - {\text{index}}(F){,_{}}d_{\text{R}}^{\text{C}} = {\text{index}}(E) - (\varepsilon + f_{}^c - 1) $, $ d_\varepsilon ^{\text{C}} = |d_{\text{R}}^{\text{C}} - d_{\text{L}}^{\text{C}}| $，其中$ \varepsilon \in (z,n) $；选取最小值$ d_\varepsilon ^{\text{C}} $对应的ε值，并将时变业务
带宽压缩后所需保护频隙数目$ {f^c} $分配在频隙索引值区间$ [\varepsilon ,\varepsilon + f_{}^c - 1] $的频谱块上，输出结果，结束算法3。
下载: 导出CSV
表 1 默认仿真参数
参数 数值 时变业务请求数目 105 个 每条链路提供的频隙数目F 320 fs 单位频隙提供带宽Cf 12.5 GHz 业务保护等级q (0, 1) 业务大小 [25, 125] Gb/s 业务请求到达率 服从泊松分布 业务请求持续时间 服从负指数分布
下载: 导出CSV
-
[1] 柴蓉, 邹飞, 刘莎, 等. 6G移动通信: 愿景、关键技术和系统架构[J]. 重庆邮电大学学报(自然科学版), 2021, 33(3): 337–347.CHAI Rong, ZOU Fei, LIU Sha, et al. 6G mobile communication: Vision, key technologies and system architecture[J]. Journal of Chongqing University of Posts and Telecommunications (Natural Science Edition), 2021, 33(3): 337–347. [2] LIU Huanlin, REN Jie, CHEN Yong, et al. Spectrum slicing-based fragmentation aware routing and spectrum allocation in elastic optical networks[J]. Optical Switching and Networking, 2022, 45: 100673. doi: 10.1016/j.osn.2022.100673 [3] 付亚伟. 大数据互联网时代光纤通信技术的发展与挑战[J]. 重庆邮电大学学报(自然科学版), 2021, 33(1): 52–58.FU Yawei. Development and challenge of optical fiber communication technology in the era of big data internet[J]. Journal of Chongqing University of Posts and Telecommunications (Natural Science Edition), 2021, 33(1): 52–58. [4] KATO M and BABA K I. A multi-path routing method with traffic grooming corresponding to path lengths in elastic optical networks[J]. IEICE Transactions on Communications, 2022, E105.B(9): 1033–1038. doi: 10.1587/transcom.2021EBP3196 [5] SHAHRIAR N, TAEB S, CHOWDHURY S R, et al. Reliable slicing of 5G transport networks with bandwidth squeezing and multi-path provisioning[J]. IEEE Transactions on Network and Service Management, 2020, 17(3): 1418–1431. doi: 10.1109/TNSM.2020.2992442 [6] BAO Bowen, YANG Hui, YAO Qiuyan, et al. SDFA: A service-driven fragmentation-aware resource allocation in elastic optical networks[J]. IEEE Transactions on Network and Service Management, 2022, 19(1): 353–365. doi: 10.1109/TNSM.2021.3116757 [7] GOŚCIEŃ R and KSIENIEWICZ P. Efficient dynamic routing in spectrally-spatially flexible optical networks based on traffic categorization and supervised learning methods[J]. Optical Switching and Networking, 2022, 43: 100650. doi: 10.1016/j.osn.2021.100650 [8] KUMAR D, KUMAR R, and SHARMA N. Proactive connection recovery strategy with recovery time constraint for survivable elastic optical networks[J]. China Communications, 2021, 18(9): 236–248. doi: 10.23919/JCC.2021.09.018 [9] DING Shifeng, BOSE S K, and SHEN Gangxiang. Spectrum trading between virtual optical networks with time-varying traffic in an elastic optical network[J]. Journal of Optical Communications and Networking, 2020, 12(3): 24–37. doi: 10.1364/JOCN.377462 [10] PATHAK S K and PRAKASH S. Traffic aggregation based RSA (TARSA) for time varying sub-wavelength connections in elastic optical networks[J]. Wireless Personal Communications, 2021, 118(4): 2733–2748. doi: 10.1007/s11277-021-08152-5 [11] CHRISTODOULOPOULOS K, TOMKOS I, and VARVARIGOS E. Time-varying spectrum allocation policies and blocking analysis in flexible optical networks[J]. IEEE Journal on Selected Areas in Communications, 2013, 31(1): 13–25. doi: 10.1109/JSAC.2013.130103 [12] SHEN Gangxiang, GUO Hong, and BOSE S K. Survivable elastic optical networks: Survey and perspective (invited)[J]. Photonic Network Communications, 2016, 31(1): 71–87. doi: 10.1007/s11107-015-0532-0 [13] TANG Fengxian, GANG Gangxiang, and ROUSKAS G N. Crosstalk-aware shared backup path protection in multi-core fiber elastic optical networks[J]. Journal of Lightwave Technology, 2021, 39(10): 3025–3036. doi: 10.1109/JLT.2021.3064935 [14] 鲍宁海, 苏国庆, 陈静波. 恢复时间敏感的光网络混合通路保护算法[J]. 重庆邮电大学学报(自然科学版), 2017, 29(3): 313–319.BAO Ninghai, SU Guoqing, and CHEN Jingbo. Recovery-time aware hybrid path protection algorithm in optical networks[J]. Journal of Chongqing University of Posts and Telecommunications (Natural Science Edition), 2017, 29(3): 313–319. [15] YIN Shan, GUO Shijia, MENG Xiangkai, et al. XT-considered multiple backup paths and resources shared protection scheme based on ring covers[J]. Optics Express, 2021, 29(5): 6737–6755. doi: 10.1364/OE.400042 [16] DING D R. Spectrum expansion/contraction and survivable routing and spectrum assignment problems on EONs with time-varying traffic[J]. Computer Communications, 2019, 148: 152–164. doi: 10.1016/j.comcom.2019.09.017 [17] PAIRA S, HALDER J, CHATTERJEE M, et al. On energy efficient survivable multipath based approaches in space division multiplexing elastic optical network: Crosstalk-aware and fragmentation-aware[J]. IEEE Access, 2020, 8: 47344–47356. doi: 10.1109/ACCESS.2020.2979487 [18] SHAHRIAR N, ZULFIQAR M, RAHMAN S, et al. Disruption minimized bandwidth scaling in EON-enabled transport network slices[J]. IEEE Journal on Selected Areas in Communications, 2021, 39(9): 2734–2747. doi: 10.1109/JSAC.2021.3064643 -
下载:
图(4) / 表(4)
计量
- 文章访问数: 431
- HTML全文浏览量: 122
- PDF下载量: 33
- 被引次数: 0
