Skip to main content
Log in

Analysis and optimization of a cross-layer adaptation mechanism for real-time applications in wireless networks

  • Published:
Wireless Networks Aims and scope Submit manuscript

Abstract

The theoretical analysis of a cross-layer mechanism for improving the quality of service of real-time applications in wireless networks is presented. The mechanism coordinates adaptations of the modulation order at the Physical layer and the media encoding mode at the Application layer, to improve packet loss rate, throughput and mean delay. With the use of Continuous Flow Modeling, the system is considered as a “fluid” queue with inflow and outflow rates representing its traffic generation and service rates, respectively. Each data source is modeled as a Markov chain, from the steady-state of which the optimal adaptation thresholds of the cross-layer mechanism are derived. Extensive performance evaluation results show that the optimized operation of the mechanism attains a significant performance improvement compared to both the sub-optimal case, and a legacy system, which adjusts the modulation order and encoding mode separately and independently of each other.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9

Similar content being viewed by others

References

  1. Vutukuru, M., Balakrishnan H., & Jamieson, K. K. (2009). Cross-layer wireless bit rate adaptation. ACM SIGCOMM Computer Communication Review, 39(4), 3–14.

    Article  Google Scholar 

  2. Choi, J., Na, J., Lim, Y., Park, K., & Kim, C. (2008). Collision-aware design of rate adaptation for multi-rate 802.11 WLANs. IEEE Journal on Selected Areas in Communications, 26(8), 1366–1375.

    Article  Google Scholar 

  3. Ge, W., Zhang, J., & Shen, S. (2007). A cross-layer design approach to multicast in wireless networks. IEEE Transactions on Wireless Communications, 6(3), 1063–1071.

    Article  Google Scholar 

  4. Liu, Q., Zhou, S., & Giannakis, G. B. (2004). Cross-Layer combining of adaptive modulation and coding with truncated ARQ over wireless links. IEEE Transactions on Wireless Communications, 3(5), 1746–1755.

    Article  Google Scholar 

  5. Hoang, A. T., & Motani, M. (2008). Cross-layer adaptive transmission with incomplete system state information. IEEE Transactions on Communications, 56(11), 1961–1971.

    Article  Google Scholar 

  6. Choudhury, S. & Gibson, J. D. (2007). Payload length and rate adaptation for multimedia communications in wireless LANs. IEEE Journal on Selected Areas in Communications, 25(4), 796–807.

    Article  Google Scholar 

  7. Djama, I., Ahmed, T., Nafaa, A., & Boutaba, R. (2008). Meet in the middle cross-layer adaptation for audiovisual content delivery. IEEE Transactions on Multimedia, 10(1), 105–120.

    Article  Google Scholar 

  8. Triantafyllopoulou, D., Passas, N., Salkintzis, A., & Kaloxylos, A. (2007). A Heuristic cross-layer mechanism for real-time traffic over IEEE 802.16 Networks. Wiley’s International Journal of Network Management, special issue on “Management Solutions for QoS Support over the Entire Audio-Visual Service Distribution Chain, 17(5), 347–361.

    Google Scholar 

  9. Wu, D., Ci, S., & Wang, H. (2007). Cross-Layer optimization for video summary transmission over wireless networks. IEEE Journal on Selected Areas in Communications, 25(4), 841–850.

    Article  Google Scholar 

  10. IEEE Std 802.16e, “IEEE standard for local and metropolitan area networks part 16: Air interface for fixed and mobile broadband wireless access systems. Amendment 2: Physical and medium access control layers for combined fixed and mobile operation in licensed bands and corrigendum 1. Feb. 2006.

  11. Khan, S., Peng, Y., Steinbach, E., Sgroi, M., & Kellerer, W. (2006). Application-driven cross-layer optimization for video streaming over wireless networks. IEEE Communications Magazine, 44(1), 122–130.

    Article  Google Scholar 

  12. Wardi, Y., & Melamed, B. (2001). Variational bounds and sensitivity analysis of traffic processes in continuous flow models. Springer’s Discrete Event Dynamic Systems, 11(3), 249–282.

    Article  MathSciNet  MATH  Google Scholar 

  13. Wardi, Y., & Melamed, B. (2000). Loss volume in continuous flow models: Fast simulation and sensitivity analysis via IPA. Proceedings of 8th IEEE Mediterranean Conference on Control and Automation (MED 2000), July 2000.

  14. Liu, B., Guo, Y., Kurose, J., Towsley, D., & Gong, W. B. (1999). Fluid simulation of large scale networks: Issues and tradeoffs. Proceedings of International Conference on Parallel and Distributed Processing Techniques and Applications, June 1999.

  15. Kumaran, K., & Mitra, D. (2001). Performance and fluid simulations of a novel shared buffer management systems. ACM Transactions on Modeling and Computer Simulation (TOMACS), 11(1), 43–75.

    Article  Google Scholar 

  16. Yan, A., & Gong, W. B. (1999). Fluid simulation for high speed networks with flow-based routing. IEEE Transactions on Information Theory, 45(5), 1588–1599.

    Article  MathSciNet  MATH  Google Scholar 

  17. Liu, Q., Zhou, S., & Giannakis, G. B. (2005). Queuing with adaptive modulation and coding over wireless links: Cross-Layer analysis and design. IEEE Transactions on Wireless Communications, 4(3), 1142–1153.

    Article  Google Scholar 

  18. Cinlar, E. (1975). Introduction to Stochastic processes. NJ: Prentice Hall Inc.

  19. Mathworks Matlab http://www.mathworks.com/.

  20. 3GPP; Technical specification group services and system aspects; mandatory speech CODEC speech processing functions; AMR speech CODEC; General Description (3GPP TS 26.071 V6.00 (2004–12)).

  21. 3GPP; Technical specification group services and system aspects; mandatory speech codec speech processing functions; Adaptive Multi-Rate (AMR) speech codec frame structure (3GPP TS 26.101 V6.00 (2004-09)).

Download references

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Dionysia Triantafyllopoulou.

Appendix

Appendix

Let the Markov chain that models the operation of the system employing CLEMA be in state \( i=\left( d_{i}, m_{i}, x_{i}\right) \), \( i \in E \). For this state, the BER thresholds based on which the probabilities of the possible transitions are derived are the following:

  1. 1.

    \( th_{\varepsilon_{max}}=1-\left( \frac{1-\varepsilon_{max}} {1-\frac{L_{i}}{\alpha_{i}\cdot T_{f}}}\right) ^{\frac{1} {b_{m_{i}}}} \), so that \( R_{loss}>\varepsilon_{max} \) if \( BER_{i}>th_{\varepsilon_{max}} \)

  2. 2.

    \( th_{\varepsilon_{min}}=1-\left( \frac{1-\varepsilon_{min}} {1-\frac{L_{i}}{\alpha_{i}\cdot T_{f}}}\right) ^{\frac{1} {b_{m_{i}}}} \), so that \( R_{loss}<\varepsilon_{min} \) if \( BER_{i}<th_{\varepsilon_{min}} \)

  3. 3.

    \( th_{\delta_{low}}=1-\left( 1-\frac{\frac{L_{i}}{\alpha-{i}\cdot T_{f}}\cdot \delta_{low}}{1-\delta_{low}\cdot\left( 1-\frac{L_{i}}{\alpha_{i}\cdot T_{f}}\right) } \right)^{\frac{1}{b_{m_{i}}}} \), so that δ > δ low if BER i  > \(th_{\delta_{low}}\)

  4. 4.

    \( th_{\delta_{med}}=1-\left( 1-\frac{\frac{L_{i}}{\alpha-{i}\cdot T_{f}}\cdot \delta_{med}}{1-\delta_{med}\cdot\left( 1-\frac{L_{i}}{\alpha_{i}\cdot T_{f}}\right) } \right)^{\frac{1}{b_{m_{i}}}} \), so that δ > δ med if BER i  > \(th_{\delta_{med}}\)

The state \( i=\left( d_{i}, m_{i}, x_{i}\right) \), \( i \in E \) is absorbing \( \left( P\left( i,i\right)=1 \right) \) if the values of the thresholds δ low , δ med and β are such that satisfy the following conditions:

Let the state i is such that 1 < m i  < M and 1 < d i  < D. This state is absorbing if all the following conditions are met:

  1. 1.

    The probability of transition to a state with lower modulation order is zero: \( P\left( \left( d_{i},m_{i},x_{i}\right) , \left( d_{i},m_{i}-1,x_{j}\right)\right) =0 \), thus if:

    $$ P\left( BER_{i}>max\left\lbrace th_{\varepsilon_{max}},th_{\delta_{med}}\right\rbrace \right) =0. $$
  2. 2.

    The probability of transition to a state with higher modulation order is zero: \( P\left( \left( d_{i},m_{i},x_{i}\right) , \left( d_{i},m_{i}+1,x_{j}\right)\right) =0 \), thus if:

    1. a.

      If the buffer load x i is such that the percentage of the mean delay with respect to the maximum tolerable delay is higher than \( \beta \, \left( x_{i}:\, \frac{\bar{S}}{S_{max}}>\beta \right) \):

      $$ P\left( BER_{i}>th_{\varepsilon_{max}},\, BER_{i}<th_{\delta_{low}}\right) =0\, \hbox{and}\, P\left( BER_{i}<th_{\varepsilon_{min}}\right) =0. $$
    2. b.

      If the buffer load x i is such that the percentage of the mean delay with respect to the maximum tolerable delay is lower than \( \beta \, \left( x_{i}:\, \frac{\bar{S}}{S_{max}}<\beta \right) \):

      $$ P\left( BER_{i}>th_{\varepsilon_{max}},\, BER_{i}<th_{\delta_{low}}\right) =0. $$
  3. 3.

    The probability of transition to a state with lower encoding mode is zero: \( P\left( \left( d_{i},m_{i},x_{i}\right) , \left( d_{i}-1,m_{i},x_{j}\right)\right) =0 \), thus if

    $$ P\left( BER_{i}>th_{\varepsilon_{max}},\, BER_{i}>th_{\delta_{low}},\, BER_{i}<th_{\delta_{med}}\right) =0. $$
  4. 4.

    The probability of transition to a state with higher encoding mode is zero: \( P\left( \left( d_{i},m_{i},x_{i}\right) , \left( d_{i}+1,m_{i},x_{j}\right)\right) =0 \), thus if

    $$ P\left( BER_{i}<th_{\varepsilon_{min}}\right) =0,\, x_{i}:\, \frac{\bar{S}}{S_{max}}<\beta. $$
  5. 5.

    The buffer load after the time interval of T f seconds equals x i (see Eq. 1).

In any other case, where either m i  = 1 or m i  = M or d i  = 1 or d i  = D, the state i is absorbing if all the above conditions are met except the ones that refer to impossible transitions (for example the condition of zero probability of a transition to a state with lower modulation order in case m i  = 1).

Due to the structure of the Markov chain and the calculation of its transition matrix, some of its states are transient. For example, although the state \( i:\, \left( d_{i}, m_{i}, 0\right) \) with \( \frac{\mu_{d_{i}}}{b_{m_{i}}}>\varphi \) can be considered as the initial state, the Markov chain cannot return to it due to the fact that the queue cannot empty when the inflow rate is higher that the outflow rate. Thus, in order to calculate the mean loss rate of the system modeled by the Markov chain, the problem is limited to its minimum closed set that is irreducible, as it has a finite number of states.

Rights and permissions

Reprints and permissions

About this article

Cite this article

Triantafyllopoulou, D., Passas, N., Zervas, E. et al. Analysis and optimization of a cross-layer adaptation mechanism for real-time applications in wireless networks. Wireless Netw 17, 1339–1354 (2011). https://doi.org/10.1007/s11276-011-0352-9

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11276-011-0352-9

Keywords

Navigation