Quantization of Neural Networks is popular technique for adopting computation intensive Deep Learning applications to edge devices. In this work, low bit mixed precision quantization of FPN-Resnet18 model trained for the task of semantic segmentation is explored using Cityscapes and Arriver datasets. The Hessian information of each layer in the model is used to determine the bit precision for each layer and in some experiments the bit precision for the layers are determined randomly. The networks are quantization-aware trained with bit combinations 2, 4 and 8. The results obtained for both Cityscapes and Arriver datasets show that the quantization-aware trained networks with the low bit mixed precision technique offer a performance at par with the 8-bit quantization-aware trained networks and the segmentation performance degrades when the network activations are quantized below 8 bits. Also, it was found that the usage of the Hessian information had little effect on the network’s performance. / Kvantisering av Neurala nätverk är populär teknik för att införa beräknings-intensiva Deep Learning -applikationer till edge-enheter. I detta arbete utforskas låg bitmixad precisionskvantisering av FPN-Resnet18-modellen som är utbildad för uppgiften för semantisk segmentering med hjälp av Cityscapes och Arriverdatauppsättningar. Hessisk information från varje lager i modellen, används för att bestämma bitprecisionen för respektive lager. I vissa experiment bestäms bitprecision för skikten slumpmässigt. Nätverken är kvantiserings medvetna utbildade med bitkombinationer 2, 4 och 8. Resultaten som erhållits för både Cityscapes och Arriver datauppsättningar visar att de kvantiserings medvetna utbildade nätverken med lågbit blandad precisionsteknik erbjuder en prestanda i nivå med 8-bitars kvantiseringsmedvetna utbildade nätverk och segmenteringens prestationsgrader när nätverksaktiveringarna kvantiseras under 8 bitar. Det visade sig också att användningen av hessisk information hade liten effekt på nätets prestanda.

The state-of-the-art object detection and image classification methods can perform impressively on more than 9k classes. In contrast, the number of classes in semantic segmentation datasets are fairly limited. This is not surprising , when the restrictions caused by the lack of labeled data and high computation demand are considered. To efficiently perform pixel-wise classification for c number of classes, segmentation models use cross-entropy loss on c-channel output for each pixel. The computational demand for such prediction turns out to be a major bottleneck for higher number of classes. The major goal of this thesis is to reduce the number of channels of the output prediction, thus allowing to perform semantic segmentation with very high number of classes. The reduction of dimension has been approached using metric learning for the semantic feature space. The metric learning provides us the mapping from pixel to embedding with minimal, still sufficient, number of dimensions. Our proposed approximation of groundtruth class probability for cross entropy loss helps the model to place the embeddings of same class pixels closer, reducing inter-class variabilty of clusters and increasing intra-class variability. The model also learns a prototype embedding for each class. In loss function, these class embeddings behave as positive and negative samples for pixel embeddings (anchor). We show that given a limited computational memory and resources, our approach can be used for training a segmentation model for any number of classes. We perform all experiments on one GPU and show that our approach performs similar and in some cases slightly better than deeplabv3+ baseline model for Cityscapes and ADE20K dataset. We also perform experiments to understand trade-offs in terms of memory usage, inference time and performance metrics. Our work helps in alleviating the problem of computational complexity, thus paving the way for image segmentation task with very high number of semantic classes. / De ledande djupa inlärningsmetoderna inom objektdetektion och bildklassificering kan hantera väl över 9000 klasser. Inom semantisk segmentering är däremot antalet klasser begränsat för vanliga dataset. Detta är inte förvånande då det behövs mycket annoterad data och beräkningskraft. För att effektivt kunna göra en pixelvis klassificering av c klasser, använder segmenteringsmetoder den s.k. korsentropin över c sannolikhets värden för varje pixel för att träna det djupa nätverket. Beräkningskomplexiteten från detta steg är den huvudsakliga flaskhalsen för att kunna öka antalet klasser. Det huvudsakliga målet av detta examensarbete är att minska antalet kanaler i prediktionen av nätverket för att kunna prediktera semantisk segmentering även vid ett mycket högt antal klasser. För att åstadkomma detta används metric learning för att träna slutrepresentationen av nätet. Metric learning metoden låter oss träna en representation med ett minimalt, men fortfarande tillräckligt antal dimensioner. Vi föreslår en approximation av korsentropin under träning som låter modellen placera representationer från samma klass närmare varandra, vilket reducerar interklassvarians och öka intraklarrvarians. Modellen lär sig en prototyprepresentation för varje klass. För inkärningskostnadsfunktionen ses dessa prototyper som positiva och negativa representationer. Vi visar att vår metod kan användas för att träna en segmenteringsmodell för ett godtyckligt antal klasser givet begränsade minnes- och beräkningsresurser. Alla experiment genomförs på en GPU. Vår metod åstadkommer liknande eller något bättre segmenteringsprestanda än den ursprungliga deeplabv3+ modellen på Cityscapes och ADE20K dataseten. Vi genomför också experiment för att analysera avvägningen mellan minnesanvändning, beräkningstid och segmenteringsprestanda. Vår metod minskar problemet med beräkningskomplexitet, vilket banar väg för segmentering av bilder med ett stort antal semantiska klasser.

<p>Syftet med den här rapporten är att beskriva ett antal olika tillvägagångssätt man kan använda sig av då man har behov av att dela in ett lokalt nätverk i olika segment och med det även kunna reglera trafikflödet mellan segmenten. De lösningar som presenteras i arbetet är inriktade mot mikro- och småföretag.Anledningen till att vi har valt att arbeta med det här området är att vi anser att det är viktigt för organisationer att har en strukturerad och segmenterad design på sitt interna datornätverk.Vi har arbetat genom att i förväg samla in information om olika tekniker som kan tänkas lösa vårt problem, och därefter testat olika scenarion med dessa tekniker. Data har samlats in efter varje genomfört scenario och sammanställts i statistisk form för att kunna avgöra vilken metod som var att föredra.Vi har testat lösningar där man segmenterar nätverket i en lager 2-switch medan man möjliggör och förhindrar trafikflöde mellan segmenten i en router. Även lösningar där man använder en lager 3-switch har testats. På så sätt kan routningen ske direkt i switchen och det blir betydligt mindre belastning i routern. Resultatet visar att då man vill segmentera ett nätverk så är det rekommenderat att man använder sig av VLAN och ACL:er och eventuellt i kombination med en brandvägg.Slutresultatet av rapporten är att en lösning med ”router on a stick” är den billigaste lösningen och troligen den som de flesta mindre företag skulle klara sig med. Vilken lösning man väljer beror dock helt på hur mycket pengar man vill lägga på sitt nätverk samt vad kraven är.</p> / <p>The purpose of this report is to describe a number of approaches that can be used when you are in need of dividing a local area network in a number of segments, and with that also be able to control how data traffic is allowed to traverse between the different segments. The solutions that are presented are focused towards micro and small companies.The reason that we have chosen to work with this matter is that we believe it is important for organizations to have a structured and segmented design of its internal computer network.We have been working by in advance collecting information about various techniques that might solve our problem, and then testing different scenarios using these techniques. Data have been collected after each tested scenario and compiled in statistical form in order to determine which method that was preferable.We have been testing solutions were you segment the network in a layer 2 switch while you allow or deny communication between the segments in a router, and also solutions were you use a layer 3 switch. In that way you can let the routing be performed in the switch, which leads to significantly lower load on the router. The result was that if you are about to segment a local area network it is recommended that you use VLAN and ACL:s, and possibly in combination with a firewall.The final result of this report is that a solution using the “router on a stick”-technique is the cheapest one, and probably the one that most small companies would get along with. However, the solution that you choose depends completely on how much money you want to spend on your network, and also what the needs are.</p>

The current health trend in Sweden has created a gym industry where the competition hardens for each day. Dominant actors fight for customers and the customers are many. The basic need of being psychically active is with each individual and therefore everyone, in theory becomes potential customers for the gyms. The gyms targets the widest of target groups and invite with slogans like workout for everyone. However, there are problem with maintaining customers who refrain from training or completely leaving the membership. The purpose of this study is to investigate how gyms present segmentation strategy creates disloyal customers who abandon the gym despite their individual needs and how gym can through segmentation measures create more profitable relationships with more loyal customers.The method used for this was based on interviews with two groups, the non-loyal customers and the gyms as well as an content analysis of the gyms own communication. The result from the study shows a number of gaps when it comes to the communicated target groups versus which resources are put on the operational work from gym. The study's conclusion is that the current segmentation strategy does not justify the stated target group, but instead shows obvious deficiencies which lead to non loyal customers to an excessive extent. Instead a number of measures with a focus on inclusion and focus are presented

Syfte: Syftet med den här uppsatsen är att öka förståelsen för hur samhällets och teknikens senaste utveckling har bidragit med nya marknadsföringskanaler och vad detta innebär för företags marknadsföringsstrategier. Med detta syfte som utgångspunkt har tre forskningsfrågor formulerats för att avgränsa forskningen än mer: Hur kan smartphones som marknadsföringskanal hjälpa företag att precisera sin marknadsföring till kunden? På vilket sätt har den nya mobila tekniken, i form av smartphones, påverkat företags segmenteringsmöjligheter? Hur kan ett företag använda sig av smartphones, som ett verktyg för att hantera kundrelationer? Metod: Uppsatsen är resultatet av en kvalitativ studie med en huvudsaklig induktiv inriktning. Primärdatainsamlingen har skett genom kvalitativa semistrukturerade intervjuer. Slutsats: Efter att ha undersökt detta ämne har vi dragit slutsatsen att valet av smartphones som kanal kan hjälpa företag att precisera sin marknadsföring genom att telefonen är tillgänglig och personlig för kunden. Därför menar vi att kunders integritet och acceptans kommer att vara avgörande faktorer och viktiga att beakta. Timing är en väsentlig del av precisionsmarknadsföringen och vi har därför presenterat två nya begrepp; företagsstyrd timing och kundstyrd timing. Smartphones har även bidragit med nya möjligheter för företag att segmentera marknaden där vi framhållit slutsatsen att plats och kontext torde vara nya segmenteringsvariabler för företag att beakta. Vår slutsats är även att smartphones är en fördelaktig kanal när det kommer till byggande och hantering av kundrelationer och vi har sett tendenser till att kunden umgås med varumärket på ett sätt som tidigare inte varit möjligt. Vi introducerar ett nytt begrepp, interaktionsintensiva kommunikationsenheter, där smartphones ses som den ledande enhetstypen idag. Nyckelord: Mobilmarknadsföring, smartphones, precisionsmarknadsföring, kundstyrd timing[1], segmentering, kontext, relevans, integritet, kundrelationer, interaktionsintensiva kommunikationsenheter[2] [1]Begreppet är av oss introducerat och utgör en del av en egengjord modell. Se s 58, kapitel 5.1.1 [2]Vi har valt att introducera detta begrepp som ett samlingsnamn för en ny typ av marknadsföringskanal. Se s. 63, kapitel 5.1.3 / Purpose: The purpose of this study is to gain understanding of how the development of the society, as well as the technology, has increased the numbers of marketing channels and how they affect companies’ marketing strategies. In order to investigate this subject we have formulated three main questions: How can smartphones as a marketing channel help companies pinpoint their communication to the customer? How do smartphones affect companies' ability to segment the market? How can a company use smartphones, as a tool to manage customer relationships? Approach: We have done a qualitative study with an inductive approach. Primary data has been collected through seven qualitative semi-structured interviews with persons from different companies. Conclusions: After investigating this subject we have come to the conclusion that using smartphones as a marketing channel can help companies to pinpoint their marketing because of the device’s availability and because that the phone is personal to the customer. We mean that customers’ acceptance and integrity is very important to take into consideration and will be of great importance in how well companies will succeed in its use of smartphones as a marketing channel. Timing is another important question and we have therefore presented two new concepts; business-driven timing and customer-driven timing. Smartphones has also contributed with new strategies when it comes to market segmentation. We suggest place and context as new market segmentation variables so that companies could be able to find their customer in right place and in right time. We have also come to the conclusion that smartphones as a marketing channel is advantageous when it comes to building relationship to the customer and we have also seen tendencies that the customer socialize with the brand in a way that they has not been done before. We have also introduced a new concept, interactive communication devices, which is a name for a new type of marketing channels and where smartphones is the leading device today. Keywords: Mobile marketing, precision marketing, customer driven timing[3], market segmentation, context, relevance, integrity, brand, customer relationship, high interactive communication devices[4] [3]This is a new concept that we have introduced and is a part of a model that we have created. See page 58, chapter 5.1.1 [4]This is a new concept that we have introduced and is a concept for a new type of marketing channel, see page 63, chapter 5.1.3.

The ABSA Klein Karoo National Arts Festival (KKNK) is one of the biggest and most popular Afrikaans arts festivals in South Africa, and since its modest beginnings in 1994, the festival has grown significantly with an estimated 85518 visitors attending the festival in 2010. The festival furthermore has a considerable economic impact on the host community of Oudtshoorn and the surrounding regions. The direct spending by festival visitors during the 2010 festival was estimated at R109.7 million, that consists of tickets bought at the festival, memorabilia, food, restaurants and beverages, transport, accommodation and entertainment. Benefits associated with these spending patterns include an increase in job opportunities for the residents, more entertainment opportunities, improvement of residents’ lifestyles and opportunities for expanding businesses to name but a few. However, the number of tickets bought for shows and productions at the KKNK has decreased drastically over the last six years, visitor numbers have decreased and this will result in a decrease in the total revenue and economic impact this festival has on the local community of Oudtshoorn. Therefore, the festival needs to be managed in a sustainable way through improving the entertainment offered at the festival. The festival guide and promotion material should be designed in such a way that they meet the needs of visitors and offers the necessary information services to create a unique festival experience for visitors. The specific services visitors need must be provided, such as shuttle services and security. The purpose of this research is to identify the key success factors (KSFs) in managing the visitors’ experience at the KKNK. To achieve this, a quantitative survey was done by distributing 500 questionnaires amongst the visitors to the festival in 2010 from the 1st to the 8th of April. A total of 443 completed questionnaires were received back. In Chapter 2, an analysis on the background and different theories of event tourism, special events, arts festivals, management and KSFs were investigated. Previous studies on the KSFs were also identified which rendered valuable and proved that different tourism operations including events and arts festivals have different KSFs. Furthermore, this analysis revealed that there exist different markets with their own individual needs and reasons for attending the arts festivals and therefore have their own expectations of what needs to be included in a unique festival experience. Chapter 3 consists of Article 1, and the main purpose of this article was to identify the various key success factors visitors to the KKNK see as important in satisfying their needs and providing a unique festival experience. A factor analysis was performed and identified the following KSFs: Safety and Personnel, Marketing and Accessibility, Venues, Accommodation and Ablutions, General aspects and Social impact, Parking and Restaurants and Shows and Stalls. The results indicated that professional staff, adequate safety and emergency services, clear indications, easy accessible venues, quality accommodation, affordable children activities, quality food providers and a variety of shows are important KSFs to provide visitors with an unforgettable festival experience. Chapter 4’s (Article 2) main purpose was to divide or segment visitors to the KKNK based on their travel motivation to attend the festival and their rating of the importance of the KSFs into separate markets. This purpose was achieved by performing a factor analysis on the travel motivations to identify the main travel motives for visitors to attend the KKNK, this was followed by a cluster analysis based on the travel motives where three clusters were identified namely Escapists, Festival Junkies and Culture seekers. The clusters festival organisers should focus their management skills and resources on are the Festival Junkies and Culture Seekers. After performing ANOVA and Chi–square tests the results showed statistically significant differences between the three clusters based on age, years attended the festival, length of stay, total spending, all the KSFs, gender and rock shows as visitors preferred type of show or production. The results therefore reveal that there are three different markets based on their travel motives who attend the KKNK, furthermore each of these three clusters has their own preferences, characteristics and ratings on which KSF will contribute in enhancing their festival experience. This was the first study of its kind in South Africa. Therefore, it contributes to the event management literature. Festival organisers can also use the results of this research to improve the KKNK’s sustainability and success by applying the KSFs and increase tickets sales for shows and productions, attract a younger market and attract visitors who spend more at the festival, by providing for the needs of specific markets and creating a unique festival experience for each visitor. / Thesis (M.Com. (Tourism))--North-West University, Potchefstroom Campus, 2012.

The ABSA Klein Karoo National Arts Festival (KKNK) is one of the biggest and most popular Afrikaans arts festivals in South Africa, and since its modest beginnings in 1994, the festival has grown significantly with an estimated 85518 visitors attending the festival in 2010. The festival furthermore has a considerable economic impact on the host community of Oudtshoorn and the surrounding regions. The direct spending by festival visitors during the 2010 festival was estimated at R109.7 million, that consists of tickets bought at the festival, memorabilia, food, restaurants and beverages, transport, accommodation and entertainment. Benefits associated with these spending patterns include an increase in job opportunities for the residents, more entertainment opportunities, improvement of residents’ lifestyles and opportunities for expanding businesses to name but a few. However, the number of tickets bought for shows and productions at the KKNK has decreased drastically over the last six years, visitor numbers have decreased and this will result in a decrease in the total revenue and economic impact this festival has on the local community of Oudtshoorn. Therefore, the festival needs to be managed in a sustainable way through improving the entertainment offered at the festival. The festival guide and promotion material should be designed in such a way that they meet the needs of visitors and offers the necessary information services to create a unique festival experience for visitors. The specific services visitors need must be provided, such as shuttle services and security. The purpose of this research is to identify the key success factors (KSFs) in managing the visitors’ experience at the KKNK. To achieve this, a quantitative survey was done by distributing 500 questionnaires amongst the visitors to the festival in 2010 from the 1st to the 8th of April. A total of 443 completed questionnaires were received back. In Chapter 2, an analysis on the background and different theories of event tourism, special events, arts festivals, management and KSFs were investigated. Previous studies on the KSFs were also identified which rendered valuable and proved that different tourism operations including events and arts festivals have different KSFs. Furthermore, this analysis revealed that there exist different markets with their own individual needs and reasons for attending the arts festivals and therefore have their own expectations of what needs to be included in a unique festival experience. Chapter 3 consists of Article 1, and the main purpose of this article was to identify the various key success factors visitors to the KKNK see as important in satisfying their needs and providing a unique festival experience. A factor analysis was performed and identified the following KSFs: Safety and Personnel, Marketing and Accessibility, Venues, Accommodation and Ablutions, General aspects and Social impact, Parking and Restaurants and Shows and Stalls. The results indicated that professional staff, adequate safety and emergency services, clear indications, easy accessible venues, quality accommodation, affordable children activities, quality food providers and a variety of shows are important KSFs to provide visitors with an unforgettable festival experience. Chapter 4’s (Article 2) main purpose was to divide or segment visitors to the KKNK based on their travel motivation to attend the festival and their rating of the importance of the KSFs into separate markets. This purpose was achieved by performing a factor analysis on the travel motivations to identify the main travel motives for visitors to attend the KKNK, this was followed by a cluster analysis based on the travel motives where three clusters were identified namely Escapists, Festival Junkies and Culture seekers. The clusters festival organisers should focus their management skills and resources on are the Festival Junkies and Culture Seekers. After performing ANOVA and Chi–square tests the results showed statistically significant differences between the three clusters based on age, years attended the festival, length of stay, total spending, all the KSFs, gender and rock shows as visitors preferred type of show or production. The results therefore reveal that there are three different markets based on their travel motives who attend the KKNK, furthermore each of these three clusters has their own preferences, characteristics and ratings on which KSF will contribute in enhancing their festival experience. This was the first study of its kind in South Africa. Therefore, it contributes to the event management literature. Festival organisers can also use the results of this research to improve the KKNK’s sustainability and success by applying the KSFs and increase tickets sales for shows and productions, attract a younger market and attract visitors who spend more at the festival, by providing for the needs of specific markets and creating a unique festival experience for each visitor. / Thesis (M.Com. (Tourism))--North-West University, Potchefstroom Campus, 2012.

The purpose of this study is to examine what characterizes brand identity and positioning regarding festivals on the current Swedish festival market. In relation to our purpose we are also going to analyze how festival attendees relate to a festivals brand identity. Our research question is as follows: What characterizes festivals brand identity and positioning on the current Swedish festival market? This study has been written with a qualitative research method and an inductive approach. Our empirical foundation is built on a rich content based on personal interviews. We interviewed in total 17 people who all have vast knowledge in our field of study. Five of our respondents have great knowledge within the festival business while the other 12 are experienced festival attendees. In our final chapter we present our conclusions that we have reached through our analysis of the empirical materials that we have weighed against theoretical framework. The conclusions we have reached are that the festivals are dependent on their core values which is a part of their brand identity, that in turn defines their positioning on the market. Festivals also need to offer added value in order to attract new customer and to maintain their current relationship to their customers.

For smaller business it is important to have good and concrete data to make decisions because there are no big margins to test and fail or go on intuition. The purpose of this study is to create an understanding of how data can be used as a decision-making tool in small business. This paper studies the data two smaller companies collect and how they use the collected data. The study has a qualitative method with interviews which have been used for collecting result. The informants were selected based on their previous knowledge and experience.The collection of data in smaller business is both from internal and external sources that complement each other. It is important to have an understanding and knowledge of the visualizations otherwise it can be misleading. Visualizations are used to find patterns, trends and other affecting factors in the data. Segmentation is used by smaller business to understand their target market, customers and which customers they should direct their attention to. Data-driven decision making uses different sources of information, visualization and segmentation. With data-driven decision making it is important to maintain the overall perspective of the business, while all parts of the business are involved in the process. / För mindre företag är det viktigt att ha bra och konkret data för att fatta beslut eftersom det inte finns så stora marginaler att testa sig fram eller gå på intuition. Syftet med studien är att skapa förståelse för hur data kan användas som ett beslutsfattande verktyg i mindre företag. Studien undersöker vilken data mindre företag samlar in samt hur de använder den data som samlas in. Studien är en kvalitativ där intervjuer använts för datainsamling. Dessa gjordes på två mindre företag där informanterna valdes ut baserat på deras tidigare kunskaper och erfarenheter.Insamlingen av mindre företagens data sker både från interna och externa källor som kompletterar varandra. Det viktigt att ha förståelse och kunskap om visualiseringarna för att de inte ska bli missvisande. Visualiseringar används för att lättare hitta mönster, trender och andra påverkande faktorer i data. Segmentering används hos mindre företag för att förstå företagets målgrupp, kunder och vilka kunder de ska rikta sin uppmärksamhet till. För datadriven beslutsfattning används olika informationskällor, visualiseringar och segmentering. Med datadriven beslutsfattning är det viktigt att behålla helhetsperspektiv över företaget samtidigt som alla delar av företaget är involverade i processen.

Syftet med den här rapporten är att beskriva ett antal olika tillvägagångssätt man kan använda sig av då man har behov av att dela in ett lokalt nätverk i olika segment och med det även kunna reglera trafikflödet mellan segmenten. De lösningar som presenteras i arbetet är inriktade mot mikro- och småföretag.Anledningen till att vi har valt att arbeta med det här området är att vi anser att det är viktigt för organisationer att har en strukturerad och segmenterad design på sitt interna datornätverk.Vi har arbetat genom att i förväg samla in information om olika tekniker som kan tänkas lösa vårt problem, och därefter testat olika scenarion med dessa tekniker. Data har samlats in efter varje genomfört scenario och sammanställts i statistisk form för att kunna avgöra vilken metod som var att föredra.Vi har testat lösningar där man segmenterar nätverket i en lager 2-switch medan man möjliggör och förhindrar trafikflöde mellan segmenten i en router. Även lösningar där man använder en lager 3-switch har testats. På så sätt kan routningen ske direkt i switchen och det blir betydligt mindre belastning i routern. Resultatet visar att då man vill segmentera ett nätverk så är det rekommenderat att man använder sig av VLAN och ACL:er och eventuellt i kombination med en brandvägg.Slutresultatet av rapporten är att en lösning med ”router on a stick” är den billigaste lösningen och troligen den som de flesta mindre företag skulle klara sig med. Vilken lösning man väljer beror dock helt på hur mycket pengar man vill lägga på sitt nätverk samt vad kraven är. / The purpose of this report is to describe a number of approaches that can be used when you are in need of dividing a local area network in a number of segments, and with that also be able to control how data traffic is allowed to traverse between the different segments. The solutions that are presented are focused towards micro and small companies.The reason that we have chosen to work with this matter is that we believe it is important for organizations to have a structured and segmented design of its internal computer network.We have been working by in advance collecting information about various techniques that might solve our problem, and then testing different scenarios using these techniques. Data have been collected after each tested scenario and compiled in statistical form in order to determine which method that was preferable.We have been testing solutions were you segment the network in a layer 2 switch while you allow or deny communication between the segments in a router, and also solutions were you use a layer 3 switch. In that way you can let the routing be performed in the switch, which leads to significantly lower load on the router. The result was that if you are about to segment a local area network it is recommended that you use VLAN and ACL:s, and possibly in combination with a firewall.The final result of this report is that a solution using the “router on a stick”-technique is the cheapest one, and probably the one that most small companies would get along with. However, the solution that you choose depends completely on how much money you want to spend on your network, and also what the needs are.