Routing Considerations in Airline Yield Management

  • Victoria C. P. Chen
  • Dirk Günther
  • Ellis L. Johnson
Part of the Applied Optimization book series (APOP, volume 79)


The airline yield management problem considers how to allocate capacity to the different fare classes so that the airline’s revenue is maximized. In the traditional airline yield management problem it is assumed that customers specify their itineraries, i.e., flight sequences. In the context of cargo airline yield management and in the context of a low fare search, customers specify only the origin and destination of their travel, a price and a time window for the travel. When the demand is origin and destination specific, the airline yield management problem not only must decide which requests to accept versus reject, as in itinerary specific demand, but must also solve the routing problem for each accepted booking request.

With regard to price-driven passengers, airlines are interested in keeping these customers away from itineraries that are preferred by those who are willing to pay higher fares. A routing algorithm that is tailored towards these objectives must take into account actual costs as well as some kind of toll representing opportunity cost. We present a fast routing algorithm that is tailored towards the special structure of airline networks to do the route generation.

For the airline yield management problem, we discuss extensions of three approaches for itinerary specific demand to handle origin and destination specific demand. Two approaches are bid price methods, one based on a deterministic network model and one based on a stochastic network model, and the third uses an approximation of the value function in a Markov decision problem formulation of the airline yield management problem.

Finally, a comparison is presented between first-come-first-serve and bid-pricing based on the deterministic network model, using a flight network extracted from the network of a major domestic carrier. The stations are connected through 48 flight legs, with alternate routes for most station pairs. The bid pricing approach results in revenue increases between 7.2% and 28.63%.


yield management routing bid pricing 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. [1]
    Chen V. C. P. Application of MARS and orthogonal arrays to inventory forecasting stochastic dynamic programs. Computational Statistics and Data Analysis, 30 317–341, 1999.zbMATHCrossRefGoogle Scholar
  2. [2]
    Chen V. C. P. Measuring the goodness of orthogonal array discretizations for stochastic programming and stochastic dynamic programming. SI AM Journal of Optimization, 12 322–344, 2001.zbMATHCrossRefGoogle Scholar
  3. [3]
    Chen V. C. P., Günther D., Johnson E. L. Solving for an optimal airline yield management policy via statistical learning. Journal of the Royal Statistical Society, Series C, accepted, 2002.Google Scholar
  4. [4]
    Chen V. C.P., Ruppert D., Shoemaker C. A. Applying experimental design and regression splines to high-dimensional continuous state stochastic dynamic programming. Operations Research, 47 38–53, 1999.MathSciNetzbMATHCrossRefGoogle Scholar
  5. [5]
    D’Sylvia E. O and D seat assignment to maximize expected revenue. Technical report, Boeing Commercial Airplane Company, unpublished, 1982.Google Scholar
  6. [6]
    Friedman J. H. Multivariate adaptive regression splines (with discussion). Annals of Statistics, 19 1–141, 1991.MathSciNetzbMATHCrossRefGoogle Scholar
  7. [7]
    Gallego G., van Ryzin G.J. Multiproduct dynamic pricing problem and its applications to network yield management. Operations Research, 45 24–41, 1997.zbMATHCrossRefGoogle Scholar
  8. [8]
    Günther D. Airline Yield Management PhD dissertation, Georgia Institute of Technology, 1998.Google Scholar
  9. [9]
    Lautenbacher C. J., Stidham S. The underlying Markov decision process in the single-leg airline yield management problem. Transportation Science, 33 136–146, 1999.zbMATHCrossRefGoogle Scholar
  10. [10]
    McGill J., van Ryzin G. J. Revenue management: Research overview and prospects. Transportation Science, 33 233–256, 1999.zbMATHCrossRefGoogle Scholar
  11. [11]
    Sedgewick R., Vitter J. S. Shortest paths in Euclidean graphs. Proceedings of the Annual Symposium on Foundations of Computer Science, 25 417–424, 1984.Google Scholar
  12. [12]
    Simpson R. W. Using network flow techniques for origin-destination seat inventory control. Technical Report, Massachusetts Insttute of Technology, Flight Transportation Laboratory Memorandum, M89–1, 1989.Google Scholar
  13. [13]
    Talluri K., van Ryzin G.J. An analysis of bid-price controls for network revenue. Management Science, 44 1577–1593, 1998.zbMATHCrossRefGoogle Scholar
  14. [14]
    Williamson E. L. Airline Network Seat Inventory Control: Methodologies and Revenue Impacts. PhD thesis, Massachusetts Institute of Technology, 1992.Google Scholar

Copyright information

© Springer Science+Business Media Dordrecht 2003

Authors and Affiliations

  • Victoria C. P. Chen
    • 1
  • Dirk Günther
    • 2
  • Ellis L. Johnson
    • 3
  1. 1.Department of Industrial and Manufacturing Systems EngineeringUniversity of Texas at ArlingtonUSA
  2. 2.Sabre Research GroupSouthlakeUSA
  3. 3.School of Industrial and Systems EngineeringGeorgia Institute of TechnologyUSA

Personalised recommendations