Team Cadel: Simulation of 4G and 3G Cell Network

Presentation at Agent Based Modelling Workshop. University of

Sydney July 28th 2011.

Overview of the Research Problem

• How do managers of networks (cell and broadband) introduce new technology?– Sale of Motorola Network see Crockett, (2008)

• New technology may be more expensive and limited in capacity.– Mah (2008)

• Consumers may move to another provider (Churn) if they are not happy with service.– Dierkes et al. (2011)

• Consumers may make decisions on price and access to new technology. – Jack, (2008) and Poynter, (2006).

Background research.

• Access to better technology.– Chang Hyun, J. and J. Villegas (2008) and Colwell, S. R., M.

Aung, et al. (2008).

• Price and Value– Deng, Z., Lu, Yaobin, Wei, Kwok Kee and Zhang, Jinlong (2010).

• Both factors– Goode, Davies, Moutinho, and Jamal, (2005). – Iyengar, et al. (2008).

The Netlogo Model: Moving from a 3G to 4G World

• Network– 4G- Number of 4G cell access points (0-100)– 3G- Remaining cell access points– Capacity of each cell (0-100).

• Price– Price of 4G (0-50)– Price 3G (0-50)

• Consumers– Tolerance (of bad service) – (0-5).

Behaviour of Agents

1. Agents are set with a random allocation of bandwidth requirement and price acceptance.

2. Agents seek to maximise bandwidth at an access point.a) Green patch 3G

b) Black patch 4G.

3. Agents seek to mimimize price of each offering.• Agents who do 1 and 2 are happy (RED) and don’t move.

• Agents who cannot find a combination of 1 and 2 are not

happy (WHITE) and after 10 turns, leave for another provider (they die). This decided by Churn which can be reduced by Tolerance.

The Netlogo Model

Experimental Design

• Independent variables• ["FourG" 5 10]• ["capacity" 5 10]• ["Price4G" 10 20]• ["Price3G" 10 20]• ["tolerance" 0 1]• ["Customers" 500]• 32 Runs

• Dependent Variables• % Happy customers• Loss of customers• Mean use of 3G• Mean use of 4G

Descriptive Results

Statistics

countturtles Happy Loss Mean3G Mean4G

N Valid 1008.00 1008.00 1008.00 1008.00 1008.00

Missing .00 .00 .00 .00 .00

Mean 495.59 46.10 4.41 .48 .52

Median 500.00 48.40 .00 .49 .43

Mode 500.00 50.20 .00 .00 .00

Std. Deviation 35.30 18.19 35.30 .25 .36

Skewness -8.11 -.34 8.11 .39 .97

Std. Error of Skewness .08 .08 .08 .08 .08

Kurtosis 65.41 .75 65.41 -.42 .39

Std. Error of Kurtosis .15 .15 .15 .15 .15

Minimum 154.00 .00 .00 .00 .00

Maximum 500.00 100.00 346.00 1.07 1.80

Percentiles 25 500.00 36.65 .00 .28 .26

50 500.00 48.40 .00 .49 .43

75 500.00 57.55 .00 .60 .73

Results

Source Dependent Variable df F Sig.

Partial Eta Squared

FourG Happy 1 53.16 0.00 0.05 Mean3G 1 188.75 0.00 0.16 Mean4G 1 881.73 0.00 0.47

Price3G Happy 1 66.24 0.00 0.06 Mean3G 1 55.29 0.00 0.05 Mean4G 1 39.68 0.00 0.04

tolerance Happy 1 118.94 0.00 0.11 Loss 1 69.57 0.00 0.07 Mean3G 1 278.10 0.00 0.22 Mean4G 1 15.46 0.00 0.02

capacity

Mean3G 1 494.02 0.00 0.34 Mean4G 1 559.66 0.00 0.36

FourG * tolerance

Mean3G 1 49.37 0.00 0.05 FourG * capacity

Mean3G 1 5.00 0.03 0.01 Mean4G 1 89.68 0.00 0.08

Price4G * Price3G

Mean4G 1 14.32 0.00 0.01 Price4G * capacity

Mean4G 1 3.83 0.05 0.00 Price3G * tolerance Happy 1 3.97 0.05 0.00 Price3G * capacity

Mean3G 1 4.47 0.03 0.00 Mean4G 1 7.97 0.00 0.01

tolerance * capacity

Mean3G 1 18.43 0.00 0.02 FourG * Price4G * Price3G Mean4G 1 9.37 0.00 0.01 FourG * Price4G * tolerance Mean4G 1 3.15 0.08 0.00

ANOVA Results continued

FourG * Price4G *

capacity Mean4G 1 4.86 0.03 0.00

Price4G * Price3G *

capacity Mean4G 1 22.49 0.00 0.02

Source

Dependent

Variable df F Sig.

Partial Eta

Squared

FourG * Price4G *

Price3G * tolerance Mean4G 1 4.59 0.03 0.00

FourG * Price4G *

Price3G * capacity Mean4G 1 11.94 0.00 0.01

FourG * Price4G *

tolerance * capacity Mean4G 1 11.10 0.00 0.01

FourG * Price3G *

tolerance * capacity Mean4G 1 4.64 0.03 0.00

Conclusions

• Tolerance, Four G Access points numbers capacity and only the price of 3G seemed to influence the happiness of consumers and the mean use of technology (4G and 3G).

• Loss of customers is determined mainly by tolerance.

• There are a number of interactions but the effect sizes are small.

Implications for Providers

• Planning for capacity is important but you don’t have to provide access to all consumers.

• The price of the old technology is more important than the price of the new technology.

• Relationship marketing is important to increase tolerance and prevent churn.

ReferencesChang Hyun, J. and J. Villegas (2008). "MOBILE PHONE USERS' BEHAVIORS: THE MOTIVATION

FACTORS OF THE MOBILE PHONE USER." International Journal of Mobile Marketing 3(2): 4-14.

Colwell, S. R., M. Aung, et al. (2008). "Toward a measure of service convenience: multiple-item scale development and empirical test." Journal of Services Marketing 22(2/3): 160-169

Crockett, R. O. (2008). "MOTOROLA SETS ITS PHONE UNIT FREE." BusinessWeek(4078): 36-38.

Deng, Z., Lu, Yaobin, Wei, Kwok Kee and Zhang, Jinlong (2010). "Understanding customer satisfaction and loyalty: An empirical study of mobile instant messages in China." International Journal of Information Management 30(4): 289-300.

Dierkes, T., Bichler, Martin and Krishnan, Ramayya (2011). "Estimating the effect of word of mouth on churn and cross-buying in the mobile phone market with Markov logic networks." Decision Support Systems 51(3): 361-371.

Goode, M. M. H., Davies, Fiona, Moutinho, Luiz and Jamal, Ahmad (2005). "Determining Customer Satisfaction From Mobile Phones: A Neural Network Approach." Journal of Marketing Management 21(7/8): 755-778.

Iyengar, R., Jedidi, Kamel and Kohli, Rajeev (2008). "A Conjoint Approach to Multipart Pricing." Journal of Marketing Research (JMR) 45(2): 195-210.

Jack, L. (2008). "Public gets taste for cut price communication." Marketing Week (01419285) 31(33): 3-3.

Mah, A. (2004). "Product Innovation Case Study: '3' - A Hutchinson Brand." Marketing Review 4(2): 157-188.

Poynter, K. (2006). "Vodafone: 'Stop the clock'." Marketing (00253650): 22-22.

Questions?