Abstract
This work presents a novel interleaving crescent broadcasting protocol for near video-on-demand service. The interleaving crescent broadcasting protocol is a trade-off among the subscriber’s access latency, maximum buffer requirement, needed subscriber’s bandwidth, and maximum disk I/O transfer rate. A longer subscriber’s access latency may cause a subscriber to leave. A lower maximum buffer requirement, a lower needed subscriber’s bandwidth, and a lower maximum disk I/O transfer rate reduce subscribers’ costs. The interleaving crescent broadcasting protocol not only makes access latency shorter, but also lowers the overall system’s cost. We prove the correctness of the interleaving crescent protocol; provide mathematical analyses to demonstrate its efficiency.
Similar content being viewed by others
References
Chang YH, Coggins D, Pitt D, Skellern D (1994) An open-system approach to video on demand. IEEE Commun Mag 32:68–80
Chen YW (2004) An enhanced recursive frequency splitting broadcasting algorithm for near video-on-demand services. Inf Process Lett 92:299–302
Chen YW, Hsieh CY (2007) An interleaving slotted-polyharmonic-staircase broadcasting protocol for near video-on-demand services. Comput Commun 30(4):807–817
Dan A, Sitaram D, Shahabuddin P (1996) Dynamic batching policies for an on-demand video server. Multimedia Syst 4:112–121
Gao L, Kurose J, Towsley D (2002) Efficient schemes for broadcasting popular videos. Multimedia Syst 8:284–294
Ghose D, Kim HJ (2000) Scheduling video streams in video-on-demand systems: a survey. Multimedia Tools and Applications 11:167–195
Guo Y, Gao L, Towsley D, Sen S (2004) Smooth workload adaptive broadcast. IEEE Trans Multimedia 6(2):387–395
Hu A (2001) Video-on-demand broadcasting protocols: a comprehensive study. IEEE INFOCOM, pp 508–517
Juhn LS, Tseng LM (1997) Staircase data broadcasting and receiving scheme for hot video service. IEEE Trans. on Consum Electron 43(4):1110–1117
Juhn LS, Tseng LM (1998) Fast data broadcasting and receiving scheme for popular video service. IEEE Trans Broadcast 44:100–105
Li ZN, Drew MS (2004) Fundamentals of multimedia. Prentice Hall, Chap. 16
Pâris JF (2001) A fixed-delay broadcasting protocol for video-on-demand. Proc. of the 10th International Conf. on Computer Communications and Networks (ICCCN’01), pp 418–423
Pâris JF, Carter SW, Long D-D (1999) A hybrid broadcasting protocol for video on demand. In Multimedia Computing and Networking, pp 317–326
Tseng YC, Yang MH, Chang CH (2002) A recursive frequency-splitting scheme for broadcasting hot videos in VOD service. IEEE Trans Commun 50(8):1348–1355
Yan E, Kameda T (2006) Generalized fibonacci broadcasting: an efficient VOD broadcasting scheme with user bandwidth limit. Discrete Appl Math 154(16):2418–2429
Yang HC, Yu HF, Tseng LM, Chen YM (2004) Interleaving staircase broadcasting and receiving scheme with loss-anticipation delivery. Proceedings of International Conference on Internet Computing, Las Vegas, Nevada, USA
Yu HF (2008) Hybrid broadcasting with small buffer demand and waiting time for video-on-demand applications. IEEE Trans Broadcast 54(2):304–311
Yu HF, Yang HC, Tseng LM (2007) Reverse fast broadcasting (RFB) for video-on-demand applications. IEEE Trans Broadcast 53(1):103–111
Yu HF, Chen YN, Yang HC, Yang ZY, Tseng LM (2008) An efficient scheme for broadcasting popular videos at low buffer demand. Comput Commun 31(10):2270–2279
Author information
Authors and Affiliations
Corresponding author
Additional information
This work was supported in part by NSC98-2221-E027-121.
Appendix A. Proof of Lemma 5
Appendix A. Proof of Lemma 5
By Theorem 4, we only consider the sum of the rates at which data is transferred from four consecutive channels to the subscriber’s disk. Let R(t) be the disk input transfer rate in the time slot t, \( size\left( {i,x} \right)\,\left( { = d \cdot b/div(n)} \right) \) be the amount of data which can be transferred within one time slot in the subsubchannel \( C\left[ n \right]\left( { = C_{{i,1}}^x } \right),\,D\left( {t,i,x} \right) \) be the amount of incoming data from subchannel \( C_i^x \) where t is a time slot, \( 3 \le i \le k \) and x∈{A,B}. Let us consider 11 cases of the value R(t) in time slots \( \beta_i^A, \gamma_{{i + 1}}^A, \beta_i^B, \gamma_{{i + 1}}^B, \alpha_{{i + 3}}^A \left( { = \alpha_{{i + 3}}^B } \right),\delta_i^B \left( { = \beta_{{i + 1}}^A - 1} \right) \) and the variation \( R\left( {t + 1} \right) - R(t) \) for each time slot, \( \beta_i^A \le t \le \beta_{{i + 1}}^A \), to show this lemma.
Case 1:
\( t = \beta_i^A . \) In this case,\( D\left( {t,i,A} \right) = D\left( {t,i,B} \right) = db/2 \), \( D\left( {t,l,x} \right) \ne 0 \) for \( l = i + 1,\,i + 2 \), and \( D\left( {t,i + 3,x} \right) = 0 \) for x∈{A,B} because \( \alpha_{{i + 3}}^A = \alpha_{{i + 3}}^B > \beta_i^A \). However, these data are evenly downloaded in time interval d. Therefore,
Case 2:
\( \beta_i^A \le t<\gamma_{{i + 1}}^A . \) In this case, we show R(t+1) is larger than R(t) by the following equation.
Case 3:
\( t = \gamma_{{i + 1}}^A . \) In this case, \( D\left( {t,i,A} \right) \ne 0 \), \( D\left( {t,i,B} \right) = D\left( {t,i + 1,A} \right) = db/2 \), \( D\left( {t,i + 1,B} \right) \ne 0 \), \( D\left( {t,i + 2,x} \right) \ne 0 \), and \( D\left( {t,i + 3,x} \right) = 0 \) for x∈{A,B} because \( \alpha_{{i + 3}}^A = \alpha_{{i + 3}}^B > \gamma_{{i + 1}}^A \). Therefore,
Case 4:
\( \gamma_{{i + 1}}^A \le t<\beta_i^B . \) In this case, we show R(t+1) is smaller than R(t) by the following equation.
By Case 2: \( \beta_i^A \le t<\gamma_{{i + 1}}^B, \;R\left( {t + 1} \right) - R(t) = b/\left( {16 \cdot 2^{{i - 3}} } \right) - size\left( {i + 1,A} \right)/d = b\left( {16 \cdot 2^{{i - 3}} } \right) - b/\left( {4 \cdot 2 \cdot 2^{{i - 3}} } \right) = - b/\left( {16 \cdot 2^{{i - 3}} } \right)<0. \)
Case 5:
\( t = \beta_i^B \). In this case, \( D\left( {t,i,A} \right) = 0 \), \( D\left( {t,i,B} \right) = D\left( {t,i + 1,A} \right) = db/2 \), \( D\left( {t,i + 1,B} \right) \ne 0 \), \( D\left( {t,i + 2,x} \right) \ne 0 \), and \( D\left( {t,i + 3,x} \right) = 0 \) for x∈{A,B} because \( \alpha_{{i + 3}}^A = \alpha_{{i + 3}}^B > \beta_i^B \). Therefore,
Case 6:
\( \beta_i^B \le t<\gamma_{{i + 1}}^B . \) In this case, we show R(t+1) is larger than R(t) by the following equation.
Case 7:
\( t = \gamma_{{i + 1}}^B \). In this case, \( D\left( {t,i,A} \right) = 0 \), \( D\left( {t,i,B} \right) \ne 0 \), \( D\left( {t,i + 1,x} \right) = db/2 \), \( D\left( {t,i + 2,x} \right) \ne 0 \), and \( D\left( {t,i + 3,x} \right) = 0 \) for x∈{A,B} because \( \alpha_{{i + 3}}^A = \alpha_{{i + 3}}^B > \gamma_{{i + 1}}^B \). Therefore,
Case 8:
\( \gamma_{{i + 1}}^B \left( { = 8 \cdot 2^{{i - 3}} - 1} \right) \le t<\alpha_{{i + 3}}^A \left( { = 8 \cdot 2^{{i - 3}} } \right) \). In this case, we show R(t+1) is smaller than R(t) by the following equation.
Case 9:
\( t = \alpha_{{i + 3}}^A \). In this case, \( D\left( {t,i,A} \right) = 0 \), \( D\left( {t,i,B} \right) \ne 0 \), \( D\left( {t,i + 1,x} \right) = db/2 \), \( D\left( {t,i + 2,x} \right) \ne 0 \) and \( D\left( {t,i + 3,x} \right) \ne 0 \) for x∈{A,B}. Therefore,
Case 10:
\( \alpha_{{i + 3}}^A \le t<\delta_i^B \). In this case, R(t+1) is smaller than R(t) since the equation
is the same as Case 8: \( \gamma_{{i + 1}}^B \le t<\alpha_{{i + 3}}^A \).
Case 11:
\( t = \delta_i^B \). In this case, \( D\left( {t,i,A} \right) = 0 \), \( D\left( {t,i,B} \right) \ne 0 \), \( D\left( {t,i + 1,x} \right) = db/2 \), \( D\left( {t,i + 2,x} \right) \ne 0 \), and \( D\left( {t,i + 3,x} \right) \ne 0 \) for x∈{A,B}. Therefore,
Up to now, we have calculated the values of R(t) in these time slots \( \beta_i^A, \gamma_{{i + 1}}^A, \beta_i^B, \gamma_{{i + 1}}^B \alpha_{{i + 3}}^A \left( { = \alpha_{{i + 3}}^B } \right),\delta_i^B \) and the variations R(t+1)–R(t) for each time slot t, \( \beta_i^A \le t<\beta_{{i + 1}}^A \) as summarized in Table 5. Thus \( 83b/48 \le R(t) \le 86b/48 \) holds for any time slot t with \( \beta_i^A \le t<\beta_{{i + 1}}^A \). □
Rights and permissions
About this article
Cite this article
Chen, YW., Uemura, J. An interleaving crescent broadcasting protocol for near video-on-demand services. Multimed Tools Appl 49, 299–321 (2010). https://doi.org/10.1007/s11042-009-0359-8
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11042-009-0359-8